(Lloyd A. Currie (Eds.) ) Detection in Analytical C (Book4You)

ACS SYMPOSIUM SERIES 361
Detection in Analytical
Chemistry
Importance, Theory, and Practice
Lloyd A. Currie, EDITOR
Developed from a symposium sponsored

by the Divisions of Analytical Chemistry,
Environmental Chemistry, and
Nuclear Chemistry and Technology
at the 191st Meeting
of the American Chemical Society,
New York, New York,
April 13-18, 1986
American Chemical Society, Washington, DC 1988
In Detection in Analytical Chemistry; Currie, L.;

ACS Symposium Series; American Chemical Society: Washington, DC, 1987.
Library of Congress Cataloging-in-Publication Data
Detection in analytical chemistry.
ACS symposium series; 361
Includes bibliographies and indexes.
1. Chemistry, AnalyticMeasurementCongresses.
I. Currie, Lloyd A., 1930- . II. American
Chemical Society. Division of Analytical Chemistry.
III. American Chemical Society. Division of
Environmental Chemistry. IV. American Chemical
Society. Division of Nuclear Chemistry and
Technology. V. American Chemica
(191st: 1986: New York, N.Y.) VI
QD75.4.M4D46 1988 543 87-30735
ISBN 0-8412-1445-X
Copyright 1988
American Chemical Society
All Rights Reserved. The appearance of the code at the bottom of the first page of each
chapter in this volume indicates the copyright owner's consent that reprographic copies of the
chapter may be made for personal or internal use or for the personal or internal use of
specific clients. This consent is given on the condition, however, that the copier pay the stated
per copy fee through the Copyright Clearance Center, Inc., 27 Congress Street, Salem, M A
01970, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law.
This consent does not extend to copying or transmission by any meansgraphic or
electronicfor any other purpose, such as for general distribution, for advertising or
promotional purposes, for creating a new collective work, for resale, or for information
storage and retrieval systems. The copying fee for each chapter is indicated in the code at
the bottom of the first page of the chapter.
The citation of trade names and/or names of manufacturers in this publication is not to be
construed as an endorsement or as approval by ACS of the commercial products or services
referenced herein; nor should the mere reference herein to any drawing, specification,
chemical process, or other data be regarded as a license or as a conveyance of any right or
permission, to the holder, reader, or any other person or corporation, to manufacture,
reproduce, use, or sell any patented invention or copyrighted work that may in any way be
related thereto. Registered names, trademarks, etc., used in this publication, even without
specific indication thereof, are not to be considered unprotected by law.
PRINTED IN THE UNITED STATES OF AMERICA

ACS Symposium Series
M. Joan Comstock, Series Editor
1987 Advisory Board

Harvey W. Blanch Vincent D. McGinniss
University of CaliforniaBerkele
Alan Elzerman W. H . Norton

Clemson University J. T. Baker Chemical Company
John W. Finley James C . Randall

Nabisco Brands, Inc. Exxon Chemical Company
Marye Anne Fox E . Reichmanis

The University of TexasAustin AT&T Bell Laboratories
Martin L . Gorbaty C . M . Roland

Exxon Research and Engineering Co. U.S. Naval Research Laboratory
Roland F. Hirsch W. D. Shults

U.S. Department of Energy Oak Ridge National Laboratory
G . Wayne Ivie Geoffrey K. Smith

USDA, Agricultural Research Service Rohm & Haas Co.
Rudolph J. Marcus Douglas B. Walters

Consultant, Computers & National Institute of
Chemistry Research Environmental Health

Foreword
The ACS SYMPOSIUM SERIES was founded in 1974 to provide a
medium for publishing symposia quickly in book form. The
format of the Series parallels that of the continuing ADVANCES
IN CHEMISTRY SERIES except that, in order to save time, the
papers are not typeset but are reproduced as they are submitted
by the authors in camera-read form Paper reviewed unde
the supervision of th
Advisory Board and are selected to maintain the integrity of the
symposia; however, verbatim reproductions of previously pub-
lished papers are not accepted. Both reviews and reports of
research are acceptable, because symposia may embrace both
types of presentation.

Preface
CHEMICAL MEASUREMENTS ARE CHARACTERIZED by three fundamental

processes: detection, identification, and quantification. The first of these
relates to the ultimate measurement capability as expressed in the detection
limit. The invitation to organize a symposium on this topic carried the
suggestion that we address the "true meaning" of detection limits. That
charge, in fact, influenced the structure of the symposium and the content
of this volume. The objectiv
from both fundamental an
in chemical measurement science. It is not intended to serve as a
compendium of the current "detection limits" for a broad range of
analytical methods.
The "meaning" of detection and detection limits has been examined in
two senses: (I) the basic scientific meaning as to exactly what is signified
by the minimum detectable quantity of a chemical substance, and how that
quantity is derived; and (2) the meaningfulness or usefulness of such
detection limits in the context of external problems, such as those affecting
society, industrial processes, or scientific research. It is timely and
appropriate for these two sides of detection to be considered together, for
the ability to detect prescribed amounts of chemical substances in the
natural environment, foods, or manufactured products can have important
effects on our economic or physical well-being. However, unless the
technical significance of detection decisions and detection limits is fully
defined, misleading or even dangerous conclusions can follow. Equally
important is mutual understanding by the lay public and the technical
community of their respective interpretations of detection.
This text consists of an overview chapter and four principal sections.
The first section addresses the issue of detection from the perspective of the
well-informed but nonscientific public, that is, the most extensive user
community of detection limits. The authors of Chapters 2 and 3, a former
member of Congress and a former congressional subcommittee staff
director, respectively, are eminently qualified to present the public view
because they have both helped to shape that view and because they respond
to the publics technological needs. The second section begins with a
tutorial chapter, followed by six contributions treating fundamental
characteristics of the chemical measurement process, which must be taken
into account to derive meaningful detection limits. Included in this section
vn

are expositions on comparative detection limits, a matter of some
importance in selecting chemical methods or in interpreting interlaboratory
data. Five chapters make up the third section of the book. Selected
examples illustrate the care and energy involved in devising methods of
extreme sensitivity. Exhaustive method characterization and attention to a
host of potential sources of error mark these contributions. These chapters,
together with several chapters in the preceding section, also treat important
areas of application.
The text concludes with two panel discussions. The first reflects the
overall focus of the book, in that it examines with some vigor "real-world"
problems and needs for meaningful detection, both in the laboratory and in
the regulatory environment. The second panel, comprising members of an
international team working on coding environmental analytical data, shares
some of the approaches and issues involved in preparing low-level chemical
data for computerized dat
distortion are quite significant in this regard because of problems of
rounding and truncation and interpretation of the meaning of detection.
The objectives of this book will have been met if improved communi-
cation on the subject of detection results. Such communication would be
beneficial not only between the public and the technical community, but
also within the technical community. As indicated in the overview chapter,
the history of detection limits in analytical chemistry has been marked by
an unfortunate degree of diversity in terminology and meaning and a lack
of attention to the probabilities of both false negatives and false positives.
At the same time, we should help the public understand that all detection
limits must allow for these two types of error, and that "zero" detection
limits cannot, in principle, be attained.
DISCLAIMER
This book was edited by Lloyd A. Currie in his private capacity. No official
support or endorsement by the National Bureau of Standards is intended or
should be inferred.
L L O Y D A. CURRIE
Gaithersburg, MD 20899
September 1, 1987
V111

Chapter 1
Detection: Overview of Historical, Societal,

and Technical Issues
Lloyd A . Currie
Center for Analytical Chemistry, National Bureau of Standards,

Gaithersburg, M D 20899
Practical societa
advances frequentl
possessing s p e c i f i e d detection c a p a b i l i t i e s with accept-
able p r o b a b i l i t i e s of false p o s i t i v e s and false nega-
tives. The first part of t h i s overview introduces the
basic concept of (chemical) detection, together with i t s
applicability to selected s o c i e t a l problems such as the
detection of natural hazards and the implementation of
c e r t a i n regulations. Basic scientific measurement issues
concerning assumptions and t h e i r validity, plus hypo-
thesis t e s t i n g and decision theory as r e l a t e d to analyte
detection are next introduced. Part two comprises a
b r i e f historical review, h i g h l i g h t i n g major contributions
to the concept and r e a l i z a t i o n of detection i n chemical
applications. The current state of the art i s then
considered. Part three i s the most extensive, as i t
seeks to expose most of the technical issues involved i n
deriving meaningful detection decisions and detection
limits, considering the o v e r a l l Chemical Measurement
Process. Those concerned p r i m a r i l y with s o c i e t a l or
historical matters may wish to pass over t h i s part.
Among the topics discussed are: systematic and model
error; non-normal random error; the s p e c i a l problem of
the blank; r e p l i c a t i o n vs Poisson variance; issues
concerning complex data evaluation, c a l i b r a t i o n , and
reporting -- including pitfalls associated with "black
box" algorithms; OC curves; power of the t - t e s t ; and
q u a l i t y . The section concludes with some new material on
d i s c r i m i n a t i o n limits, lower and upper regulatory l i m i t s ,
multiple detection decisions, and univariate and m u l t i -
variate identification. A brief summary follows,
bringing together historical, s o c i e t a l , and technical
highlights. A concluding observation i s that a
meaningful approach to p r a c t i c a l s o c i e t a l needs i s at
hand, but that order must be brought out of the extant
d i v e r s i t y of technical views on detection.
T h i s chapter not subject to U . S . copyright

Published 1988 A m e r i c a n C h e m i c a l Society

2 DETECTION IN ANALYTICAL CHEMISTRY
The DETECTION LIMIT (Lp) i s one o f t h e most important

c h a r a c t e r i s t i c s o f any Measurement P r o c e s s . Recognizing the
e x i s t e n c e o f such l i m i t s i s c r u c i a l b o t h f o r s t r i c t l y s c i e n t i f i c
endeavors, such as t h e s e a r c h f o r a new fundamental p a r t i c l e ( i l ,
and f o r v i t a l s o c i e t a l a p p l i c a t i o n s o f s c i e n t i f i c measurements,
such as t h e d e t e c t i o n o f a p a t h o l o g i c a l s t a t e o r a hazardous l e v e l
o f a heavy m e t a l . I n t h i s l a t t e r r e g a r d , i m p o r t a n t p r o g r e s s has
been made i n c o n v e y i n g t o t h e p u b l i c and t h e i r p o l i c y makers t h a t
i t i s a l a w o f measurement s c i e n c e t h a t t h e d e t e c t i o n c a p a b i l i t y o f
a l l Measurement P r o c e s s e s must s t o p s h o r t o f z e r o , i n c l o s e analogy
w i t h t h e T h i r d Law o f Thermodynamics.
R e c o g n i t i o n t h a t Lq may n o t be z e r o , has a l l e v i a t e d e a r l i e r
l e g i s l a t i v e problems, such as t h e d i c t u m t h a t no r e s i d u e o f p r o v e n
a n i m a l c a r c i n o g e n s may be p r e s e n t i n c e r t a i n f o o d p r o d u c t s ( 2 ) .
The f a c t , however, t h a t d e t e c t i o n l i m i t s c a n , a t a c o s t and w i t h
t e c h n o l o g i c a l advances be made e v e r s m a l l e r h a s f o r c e d reexami
nation o f regulatory issue
potential detection capabilities
c o n s i d e r a t i o n o f c o s t / b e n e f i t o r "acceptable r i s k " a l t e r n a t i v e s t o
"no d e t e c t a b l e r e s i d u e " r e g u l a t o r y p o l i c y ( 3 ) . Such a l t e r n a t i v e s
are mandatory i n l i g h t o f t h e fundamental p r i n c i p l e s o f d e t e c t i o n .
D e f i n i n g a c c e p t a b l e l e v e l s o f r i s k ( 4 ) , whether i n a r e g u l a t o r y
s e t t i n g o r w i t h r e s p e c t t o m e d i c a l d e c i s i o n s o r even i n terms o f
governmental actions i n connection with potential natural
d i s a s t e r s , i s p r i m a r i l y a s o c i o p o l i t i c a l matter. Although t h i s
i s s u e i s o f c e n t r a l importance, i t t r a n s c e n d s t h e theme o f t h i s
c h a p t e r , w h i c h i s t o examine t h e h i s t o r i c a l e v o l u t i o n and c u r r e n t
s t a t e o f t h e a r t o f d e t e c t i o n from t h e p e r s p e c t i v e o f c h e m i c a l
measurement s c i e n c e .
I n o r d e r t o h i g h l i g h t t h e importance o f D e t e c t i o n D e c i s i o n s
and Detection L i m i t s , and t o u n d e r l i n e the f a c t that the
p r o b a b i l i t y o f d e t e c t i o n does n o t i m m e d i a t e l y pass from z e r o t o
u n i t y a t t h e D e t e c t i o n L i m i t , we have p r e s e n t e d i n F i g . 1 s e v e r a l
s i t u a t i o n s where v a l i d d e t e c t i o n d e c i s i o n s and adequate d e t e c t i o n
l i m i t s a r e o f c o n s i d e r a b l e p r a c t i c a l importance. (The p r e s e n c e o f
a f i n i t e r i s k o f e r r o r ( f a l s e negative) a t the detection l i m i t
i . e . , t h e absence o f " c e r t a i n t y " i s t h e second a s p e c t o f t h e
problem t h a t i s somewhat f o r e i g n t o t h e common u n d e r s t a n d i n g , t h e
f i r s t being the f a c t that zero d e t e c t i o n l i m i t s a r e unattainable.)
This f i g u r e introduces the Hypothesis Testing foundation f o r
D e t e c t i o n , and i t demonstrates t h a t i t i s e s s e n t i a l f o r those o f us
i n v o l v e d i n measurement s c i e n c e t o d e v e l o p a sound, common, and
q u a n t i t a t i v e approach t o t h e f o r m u l a t i o n o f D e t e c t i o n L i m i t s . I n
a d d i t i o n , t h i s f o r m u l a t i o n must be communicated i n an e f f e c t i v e
manner b o t h w i t h i n t h e s c i e n t i f i c community and w i t h t h o s e who
depend on o u r measurements f o r s o c i e t a l d e c i s i o n s and p o l i c y
making.
As a f i n a l i n t r o d u c t o r y n o t e , i t s h o u l d be o b s e r v e d t h a t from
the p e r s p e c t i v e s o f b a s i c d i s c o v e r i e s i n S c i e n c e and t h e e a r l y
d i s c e r n m e n t o f fundamental changes i n t h e G l o b a l Environment (e.g.,

1. CUtfRIE Overview of Historical, Societal, and Technical Issues
H Y P O T H E S E S
CO
<l-a)
o false negative
(a)
w false positive
Q
Clouds Aircraft
Background Tremors Earthquake
Ordinary Luggage Explosives
Method B l a n k Toxic Chemical
Background R a d i a t i o n Chernobyl
Medication A b u s i v e Drug
Volume F l u c t u a t i o n s S t o r a g e Tank Leakage
F i g . 1. H y p o t h e s i s T e s t i n g and D e t e c t i o n L i m i t s . The upper

p a r t o f t h e f i g u r e i n d i c a t e s t h e n u l l [H ] and a l t e r n a t i v e [H ]
Q A
hypotheses, w i t h the corresponding d e c i s i o n s [D , D ] a t the l e f t . Q A
Two k i n d s o f erroneous d e c i s i o n s may be made: f a l s e p o s i t i v e s

[ p r o b a b i l i t y a] and f a l s e n e g a t i v e s [ p r o b a b i l i t y p]. (S r e p r e s e n t s
a s i g n a l l e v e l ; C, a d e c i s i o n p o i n t o r " c r i t i c a l " l e v e l . ) The
l o w e r s e c t i o n c o n t r a s t s a number o f " r e a l w o r l d " H 's and H 's Q A
where adequate d e t e c t i o n l i m i t s f o r t h e H 's have c l e a r , p r a c t i c a l

A
consequences.

s t r a t o s p h e r i c ozone changes, C 0 - i n d u c e d g l o b a l warming), t h e

2
a b i l i t y t o d e s i g n measurement p r o c e s s e s h a v i n g s u f f i c i e n t d e t e c t i o n
c a p a b i l i t y p l a c e s one a t t h e " c u t t i n g edge." Repeatedly i n
S c i e n c e , one f i n d s t h a t d i s c o v e r i e s a r e made j u s t as t h e s i g n a l s
b e g i n t o emerge from t h e n o i s e ; and i t i s t h e " t r a i n e d eye" w h i c h
i s g e n e r a l l y t h e f i r s t t o grasp them. Also, i n the context of
e x p e r i m e n t a l d e s i g n , i t s h o u l d be n o t e d t h a t absolute detection
l i m i t s a r e o f t e n t h e g o a l , i n t h a t t h e h y p o t h e s i s ( o r phenomenon)
t o be d e t e c t e d i s g e n e r a l l y c o n c e i v e d o f i n a b s o l u t e r a t h e r than
relative units.
1. BREADTH: The Scope o f D e t e c t i o n
F i r e s , earthquakes and o t h e r n a t u r a l h a z a r d s , p a t h o l o g i c a l s t a t e s ,
chemical contaminants, new fundamental p a r t i c l e s o r t h e o r i e s ,
instigators o r sources of pollution or crime natural or
a n t h r o p o g e n i c events o
w h e r e i n t h e b a s i c concep
the S t a t i s t i c a l Theory o f H y p o t h e s i s T e s t i n g , o c c u p i e s a c e n t r a l
position. Hypothesis formation i . e . , s p e c i f i c a t i o n o f the
source o r system s t a t e o r phenomenon t o be t e s t e d [ t h e " n u l l
h y p o t h e s i s " ] -- i s n e c e s s a r i l y t h e f i r s t s t e p . F o r example, one
might w i s h t o t e s t t h e n u l l h y p o t h e s i s ( H ) t h a t no earthquake
Q
o c c u r r e d ( a t a g i v e n time and p l a c e ) . To t e s t H , one r e q u i r e s a

Q
t e s t - o r measurement-process (MP), o f t e n a Chemical Measurement

P r o c e s s (CMP), t h e outcome o f w h i c h y i e l d s a D e c i s i o n r e g a r d i n g t h e
v a l i d i t y o f t h e n u l l h y p o t h e s i s . The " a l t e r n a t i v e h y p o t h e s i s " ( H )
A
w h i c h we w i s h t o be a b l e t o d e t e c t -- e.g., an earthquake o f a
g i v e n magnitude -- must exceed t h e D e t e c t i o n L i m i t o f t h e
Measurement P r o c e s s employed.
The keys t o u n d e r s t a n d i n g t h e meaning o f D e t e c t i o n D e c i s i o n s
and D e t e c t i o n L i m i t s i n m a t t e r s o f p r a c t i c a l importance t o s c i e n c e
and society are: a) t h e e x i s t e n c e o f t h e two s t a t e s [ o r
h y p o t h e s e s ] w h i c h we w i s h t o d i s t i n g u i s h ; b) a s p e c i f i e d measure-
ment p r o c e s s h a v i n g an adequate DETECTION LIMIT; and c ) a t h r e s h o l d
o r CRITICAL LEVEL f o r t h e measurement v a r i a b l e [ S i g n a l ] f o r making
the D e t e c t i o n D e c i s i o n . U n f o r t u n a t e l y , no measurement p r o c e s s c a n
be e x a c t , so f a l s e p o s i t i v e s [ a - e r r o r , e.g., earthquake e r r o n e o u s l y
" d e t e c t e d " ] and f a l s e n e g a t i v e s [ ^ - e r r o r , e.g., a c t u a l earthquake
missed] w i l l o c c u r . Perhaps a more common example i s t h a t o f t h e
f i r e alarm. The measurement i n t h i s case might be made w i t h a
smoke d e t e c t o r , w h i c h i f s e t t o t o o low a t h r e s h o l d might g i v e a
f a l s e a l a r m [ a - e r r o r ] due t o c o o k i n g fumes; i f t h e c r i t i c a l l e v e l
o r t h r e s h o l d i s s e t t o o h i g h , a r e a l f i r e o f some consequence might
be m i s s e d [/3-error] . I f an adequate b a l a n c e between these two
types o f e r r o r cannot be a c h i e v e d , one needs a b e t t e r measurement
p r o c e s s -- i . e . , a d e t e c t o r h a v i n g a lower d e t e c t i o n l i m i t . Note
t h a t t h e d e t e c t i o n l i m i t i s an i n h e r e n t p r o p e r t y o f t h e measurement
p r o c e s s , whereas t h e d e t e c t i o n d e c i s i o n i s made by comparing an
outcome o r r e s u l t o f measurement w i t h t h e C r i t i c a l L e v e l [ t h r e s h o l d
setting].
F i g . 1 suggests a wide range o f s i t u a t i o n s where adequate
d e t e c t i o n l i m i t s are c r u c i a l f o r the w e l l - b e i n g o f society. The
f i g u r e i m p l i e s t h a t t h e a l t e r n a t i v e h y p o t h e s i s has a unique v a l u e
on t h e x - a x i s . T h i s i s sometimes t r u e . F o r example, t h e

1. CURRIE Overview of Historical, Societal, and Technical Issues 5
radiocarbon concentration [isotope r a t i o ] f o r l i v i n g matter i s

1 A 1 2 1 2
C / C = 1 . 1 8 x l 0 ~ , w h i l s t that f o r f o s s i l f u e l carbon i s
e f f e c t i v e l y zero (5). A s i m i l a r s i t u a t i o n obtains f o r population
means for chemical concentrations indicative of certain
p a t h o l o g i c a l s t a t e s (e.g., g l u c o s e i n d i a b e t e s ( 6 ) ) , or t r a c e
element c o n c e n t r a t i o n s c h a r a c t e r i s t i c o f c e r t a i n ore b o d i e s . In a
g r e a t m a j o r i t y o f c a s e s , however, the i n t e n s i t y or magnitude
v a r i a b l e ( x - a x i s ) can take on many d i s c r e t e (denumerable H ' s ) , or A
even c o n t i n u o u s v a l u e s ( i n f i n i t e number o f H ' s ) . A Such i s the

c a s e , f o r example, w i t h c h e m i c a l or r a d i o a c t i v i t y contamination,
earthquakes, f i r e s , hurricanes, e t c . For a g i v e n measurement
p r o c e s s a s p e c i a l r e l a t i o n e x i s t s among the " d i s t a n c e " between H A
and H , and the two k i n d s o f e r r o r , a and p.

Q F i x i n g any two o f
t h e s e q u a n t i t i e s determines the t h i r d , as w i l l be shown i n a l a t e r
d i s c u s s i o n o f " O p e r a t i n g C h a r a c t e r i s t i c s . " ( S e c t i o n 3.2.3.)
1.1 Regulatory Limit

s i g n i f i c a n c e of detectio
w i t h a s p e c i f i c e x t e r n a l problem. Thus, b a s e d on q u a n t i t a t i v e
assessment o f h e a l t h e f f e c t s or o f a new s c i e n t i f i c phenomenon, one
may c o n c l u d e t h a t i t i s v i t a l t o be a b l e t o d e t e c t a s i g n a l or
c o n c e n t r a t i o n l e v e l as low as, say 1^. I t f o l l o w s t h a t a measure-
ment p r o c e s s h a v i n g 1^ no g r e a t e r t h a n 1^ must be s e l e c t e d o r ,
c o s t s p e r m i t t i n g , d e s i g n e d to meet the need. This i s i l l u s t r a t e d
i n F i g . 2 w h i c h d e p i c t s the c r i t i c a l l e v e l and d e t e c t i o n l i m i t
s c h e m a t i c a l l y f o r earthquakes. The upper p a r t o f the f i g u r e
p r e s e n t s an h y p o t h e t i c a l r e l a t i o n between damage o r s o c i e t a l c o s t
and u n d e t e c t e d earthquake magnitude, t o g e t h e r w i t h a maximum
acceptable cost which f i x e s a "regulatory l i m i t , " . (L^ might be
d e f i n e d , f o r example, by the "balance p o i n t " a t w h i c h the f a l s e
positive [ f a l s e alarm] c o s t -- the cost of evacuation, is
e q u i v a l e n t t o the f a l s e n e g a t i v e c o s t -- damage i n c u r r e d or l i v e s
l o s t i n the absence o f e v a c u a t i o n . ) The lower p a r t o f the f i g u r e
i n d i c a t e s the s i g n a l d e t e c t i o n l i m i t o f a measurement p r o c e s s w h i c h
meets t h i s need. A l s o shown i s the dependence o f Lp and the two
t y p e s o f h y p o t h e s i s t e s t i n g e r r o r s on random measurement e r r o r .
(The l o w e r p o r t i o n o f the f i g u r e , f o r a c t u a l earthquake f o r e c a s t -
i n g , r e l a t e s t o p r e c u r s o r measurement p r o c e s s e s . The w e a l t h o f
p h y s i c a l and c h e m i c a l p r e c u r s o r s u t i l i z e d are r e v i e w e d by K. Mogi
i n S c i e n c e . 1986, 233, 324.)
Two observations, perhaps o b v i o u s , f o l l o w from F i g . 2:
f i r s t , a z e r o magnitude earthquake c o u l d not i n p r i n c i p l e be made
d e t e c t a b l e ; second, w i t h i m p r o v i n g performance [ d e c r e a s e d d e t e c t i o n
l i m i t ] formerly undetectable tremors w i l l be found. Lack o f
a p p r e c i a t i o n o f t h e s e fundamental p r i n c i p l e s o f measurement may
l e a d t o r e g u l a t o r y d i f f i c u l t i e s , such as the r e q u i r e m e n t t h a t any
non-zero q u a n t i t i e s o f c h e m i c a l c a r c i n o g e n s s h o u l d be d e t e c t a b l e ,
or t h a t any d e t e c t a b l e amounts s h o u l d be r e p o r t e d ( 2 ) . The l a t t e r
has i n e f f e c t been e q u i v a l e n t to a moving t a r g e t , as a n a l y t i c a l
p r o c e d u r e s c o n t i n u e to advance d r a m a t i c a l l y [Note 11.
A f o o t n o t e on the m a t t e r o f r e g u l a t i o n , w h i c h l e a d s d i r e c t l y
t o our n e x t t o p i c , r e l a t e s to the r e l a t i v e l y r e c e n t c o s t / b e n e f i t
basis for regulatory decisions (3) and the emergence o f the
d i s c i p l i n e o f R i s k Assessment ( 4 ) . That i s , t h a t d e s p i t e the l a c k
o f any e x p l i c i t i n c o r p o r a t i o n o f a d o l l a r v a l u e on human l i f e i n

MS
(B) L C L D
F i g . 2. R e g u l a t o r y L e v e l s [L^] and D e t e c t i o n L i m i t s [ 1 ^ ] .
The upper p o r t i o n o f the f i g u r e t r a c e s a presumed r e l a t i o n between
earthquake magnitude [ a b s c i s s a ] and c o s t t o s o c i e t y [ o r d i n a t e ] .
The "Delaney amendment" v i e w p o i n t ( n o t d e f i n e d f o r e a r t h q u a k e s )
might be i n t e r p r e t e d as r e q u i r i n g z e r o s o c i e t a l r i s k and a c o r -
responding magnitude o f z e r o , which o f c o u r s e i s s c i e n t i f i c a l l y
u n a t t a i n a b l e . R a t h e r , an a c c e p t a b l e c o s t t o s o c i e t y f o r u n d e t e c t e d
e a r t h q u a k e s , h e r e imagined t o be 0.1 M}, i s used t o e s t a b l i s h the
requisite "regulatory" l e v e l . The lower p a r t o f the f i g u r e
r e p r e s e n t s the c o r r e s p o n d i n g earthquake measurement p r o c e s s o r
p r e c u r s o r a l a r m (seismograph s i g n a l , radon emanations, b i o l o g i c a l
[animal] sensors, e t c . ) . The r e q u i s i t e DETECTION LIMIT [Lp] must
now be no g r e a t e r t h a n , and Lp i n t u r n i s r e l a t e d t o the
p r o b a b i l i t y d e n s i t y f u n c t i o n s [pdf] f o r the n u l l s i g n a l [ H : S-0]Q
and the s i g n a l t o be d e t e c t e d [ H : S-Lj, ] , and a c c e p t a b l e f a l s e

A
d e c i s i o n p r o b a b i l i t i e s a, p. 1^ i s f i x e d by the H - p d f and a; Lp
Q
i s t h e n s e t by 1^ and p, g i v e n the H - p d f . A

the a l g e b r a o f r e g u l a t i o n , a de f a c t o "$2 m i l l i o n u n w r i t t e n r u l e "

has e v o l v e d ( 7 ) . An a n a l y s i s o f 10 y e a r s o f r e g u l a t o r y d e c i s i o n s
i n the US r e l a t i n g t o c h e m i c a l c a r c i n o g e n s showed t h a t t h i s v a l u e
f a i r l y c o n s i s t e n t l y marked the p o i n t above w h i c h r e g u l a t i o n s were
c l a s s i f i e d as too c o s t l y t o impose, and below w h i c h r e g u l a t i o n s
were j u d g e d as w a r r a n t e d .
Our use o f the symbol 1^, i n c i d e n t a l l y , i s n o t r e s t r i c t e d t o
r e g u l a t o r y matters. E a r t h q u a k e s , f o r example, cannot be r e g u l a t e d !
R a t h e r , 1^ denotes the e x t e r n a l l i m i t w h i c h d r i v e s the d e s i g n o f
our measurement p r o c e s s . I t c o u l d a p p l y as w e l l t o the r e q u i r e -
ments o f a h i g h q u a l i t y p r o d u c t i o n p r o c e s s , o r a t r a c e r s t u d y o f
l o n g range a t m o s p h e r i c t r a n s p o r t , o r the i n v e s t i g a t i o n o f e x t r e m e l y
slow r e a c t i o n p r o c e s s e s , or i n f a c t any o f the s i t u a t i o n s i n d i c a t e d
i n F i g . 1. Perhaps i t might b e t t e r be l a b e l e d " r e f e r e n c e l i m i t ( o r
l e v e l ) " or " r e q u i s i t e l i m i t . "
1.2 D e c i s i o n Theory an
i n t r o d u c t i o n to d e t e c t i o
P e a r s o n o r " f r e q u e n t i s t " approach t o s i g n i f i c a n c e t e s t i n g and
s i g n a l d e t e c t i o n ( 8 . 9 ) , w i t h the e x c e p t i o n o f the i m p o s i t i o n o f an
e x t e r n a l reference or r e g u l a t o r y l i m i t , , based on s o c i o p o l i t i c a l
and/or scientific considerations. An a l t e r n a t i v e approach,
e s p e c i a l l y appropriate for (detection) decisions culminating i n
some k i n d o f a c t i o n , i s p r o v i d e d by the a p p l i c a t i o n o f D e c i s i o n
Theory ( 1 0 ) , o r more g e n e r a l l y D e c i s i o n A n a l y s i s ( 1 1 ) . Although
t h i s t h e o r y may be o f c o n s i d e r a b l e importance f o r c e r t a i n s o c i e t a l
or business decision-making, i t s s t r u c t u r e i s such t h a t i t i s not
g e n e r a l l y a p p l i e d t o c h e m i c a l measurements.
The major advantages o f the d e c i s i o n t h e o r e t i c approach are
t h a t i t p e r m i t s one t o a p p l y e x p l i c i t l o s s f u n c t i o n s t o the
erroneous d e c i s i o n s [a,)9-errors] , and t h a t i t r e a d i l y i n c o r p o r a t e s
p r i o r ( o r " s u b j e c t i v e " ) knowledge c o n c e r n i n g the p r o b a b i l i t i e s o f
the r e s p e c t i v e h y p o t h e s e s . The a b i l i t y t o u t i l i z e l o s s f u n c t i o n s
and p r i o r p r o b a b i l i t y i s advantageous i n t h a t c o s t s and b e l i e f s and
v a l u e s e x t e r n a l t o the measurement p r o c e s s may be effectively
i n c o r p o r a t e d i n t o the d e c i s i o n making. A complication i s that
t h e r e may n o t be u n a n i m i t y c o n c e r n i n g the w e i g h t s t o be a s s i g n e d t o
t h e s e q u a n t i t i e s ; t h i s i s somewhat analogous t o the c o m p l i c a t i o n s
i n r e a c h i n g agreement on a p p r o p r i a t e v a l u e s f o r . [Costs, f o r
example, would d o u b t l e s s be v i e w e d d i f f e r e n t l y by r e g u l a t o r s and
r e g u l a t e e s , p r o d u c e r s and consumers, p h y s i c i a n s and p a t i e n t s , e t c .
The i s s u e i s analogous t o the q u e s t i o n o f "whose e x p e r t s " are
speaking i n C o u r t o r a d v i s i n g i n Congress -- i.e., i t is
n e c e s s a r i l y tempered by advocacy p o s i t i o n s . ] Except when one i s
t r e a t i n g a s t r i c t l y s c i e n t i f i c q u e s t i o n , however, i t i s i m p o r t a n t
t o r e a l i z e t h a t the l o s s e s and p r i o r p r o b a b i l i t i e s a r e f r e q u e n t l y
complex s o c i o p o l i t i c a l and/or economic m a t t e r s , b e s t d e t e r m i n e d by
e x p e r t s i n those f i e l d s .
D e c i s i o n theory o p e r a t e s on the b a s i s o f an "objective
f u n c t i o n " w h i c h i s i n some way o p t i m i z e d t h r o u g h the s e t t i n g o f a
d e c i s i o n threshold. A l u c i d presentation to a l t e r n a t i v e s t r a t e g i e s
f o r f o r m u l a t i n g d e t e c t i o n d e c i s i o n s has been g i v e n by L i t e a n u and
R i c a (8, p. 192). The essence o f the m a t t e r i s t h a t a t h r e s h o l d
value k Q f o r the L i k e l i h o o d R a t i o i s d e r i v e d from a) prior
p r o b a b i l i t i e s f o r the n u l l and a l t e r n a t i v e h y p o t h e s e s , b) a c o s t o r

l o s s m a t r i x s p e c i f y i n g c o s t s a s s o c i a t e d w i t h c o r r e c t and erroneous
decisions, and the p r o b a b i l i t y density functions (pdf) for
experimental outcomes f o r each o f the hypotheses i n q u e s t i o n .
These d a t a are combined t o compute the mean l o s s ( o r c o s t o r r i s k )
w h i c h i s t h e n m i n i m i z e d i n o r d e r t o d e r i v e k . The d e c i s i o n t e s t
o
i s p e r f o r m e d by comparing the o b s e r v e d ( e x p e r i m e n t a l ) v a l u e f o r k
w i t h k . [k, the l i k e l i h o o d r a t i o , i s the r a t i o o f the pdf f o r H
Q A
to that f o r H Q a t the s i g n a l l e v e l i n q u e s t i o n . ] The optimal

v a l u e , k based on the "Bayes C r i t e r i o n " i s g i v e n by the p r o d u c t o f
Q
the n e t c o s t o f a f a l s e p o s i t i v e and the p r i o r p r o b a b i l i t y o f H Q
d i v i d e d by the p r o d u c t o f the n e t c o s t o f a f a l s e n e g a t i v e and the

p r i o r p r o b a b i l i t y of H . A An i n t e r e s t i n g i l l u s t r a t i o n l e a d i n g t o
the same c o n c l u s i o n i s g i v e n i n M a s s a r t , D i j k s t r a and Kaufman (12,
p. 516) i n c o n n e c t i o n w i t h m e d i c a l d i a g n o s e s and s e l e c t i o n o f the
o p t i m a l p o i n t on the R e c e i v e r O p e r a t i n g C h a r a c t e r i s t i c Curve [ROC].
[ O p e r a t i n g C h a r a c t e r i s t i c (OC) and ROC c u r v e s w i l l be discussed
b r i e f l y i n a subsequen
"optimal" decision strateg
b o t h w e l l and i l l p a t i e n t s i s an i n t e r e s t i n g one. I n the i l l u s t r a -
t i o n p r e s e n t e d i n Ref. 12 (pp. 508 f f ) , f o r example, t h e r e i s a
presumed preponderance o f h e a l t h y p a t i e n t s [ p r i o r d i s t r i b u t i o n s ] .
By u s i n g the d i s t r i b u t i o n a l c r o s s i n g p o i n t as the t h r e s h o l d , one
f i n d s t h a t about h a l f o f the abnormal ( i l l ) s u b p o p u l a t i o n would
have been m i s d i a g n o s e d ! (See a l s o Appendix H i n Egan (9) f o r an
i n t e r e s t i n g i l l u s t r a t i o n o f the B a y e s i a n approach t o medical
d e c i s i o n making, and the consequent need f o r m u l t i p l e d i a g n o s t i c
t e s t s -- a n o n - t r i v i a l i s s u e i n the l i g h t o f c u r r e n t e f f o r t s o f
major m e d i c a l i n s u r e r s t o c u r t a i l the number o f d i a g n o s t i c t e s t s . )
R e f e r e n c e s (8) and (10) g i v e a l t e r n a t i v e d e c i s i o n s t r a t e g i e s
-- Minimax, I d e a l Observer, and Maximal L i k e l i h o o d -- when o n l y
p a r t i a l i n f o r m a t i o n i s a v a i l a b l e f o r p r i o r p r o b a b i l i t i e s and/or
costs. [The Minimax approach, f o r example, c u t s one's l o s s e s from
a wrong guess f o r the p r i o r p r o b a b i l i t i e s . ] The e f f e c t s o f s p e c i a l
p r e f e r e n c e s o r a v e r s i o n s [e.g., t o extreme c o s t ] are d i s c u s s e d i n
terms o f " U t i l i t y Theory" by Howard (11) , as w e l l as the use o f
D e c i s i o n A n a l y s i s f o r d e s i g n i n g s e q u e n t i a l e x p e r i m e n t s and the
s e t t i n g of research p r i o r i t i e s .
T h i s b r i e f e x c u r s i o n i n t o D e c i s i o n Theory i s i n c l u d e d t o
i n d i c a t e the manner i n w h i c h e x p e r i m e n t a l d a t a can be c o u p l e d w i t h
e x t e r n a l ( s o c i e t a l ) judgments t o form a l o g i c a l b a s i s f o r s o c i e t a l
d e c i s i o n s and a c t i o n s . A j u s t i f i c a t i o n f o r so complex a s t r a t e g y
f o r d e c i s i o n making i s t h a t " s i m p l e " s c i e n t i f i c measurements and
model e v a l u a t i o n s w i l l always be c h a r a c t e r i z e d by measurement
uncertainty. Y e t s o c i e t a l d e c i s i o n s and a c t i o n s must take p l a c e
even under the shadow o f u n c e r t a i n t y . For s c i e n t i f i c measurements,
as d i s c u s s e d i n the f o l l o w i n g t e x t , however, we s h a l l r e s t r i c t our
attention to the relatively simple Neyman-Pearson hypothesis
t e s t i n g model (8, p. 198).
1.3 Testing of Assumptions. The detection of erroneous

assumptions l i e s a t the core o f sound measurement s c i e n c e . I t i s
t h e r e f o r e e s p e c i a l l y appropriate to i n c l u d e reference to D e t e c t i o n
D e c i s i o n s and D e t e c t i o n L i m i t s f o r key assumptions i n our s u r v e y o f
the scope o f D e t e c t i o n . Assumptions o f p r i n c i p a l importance f o r
c h e m i c a l measurements i n c l u d e those r e l a t i n g t o the f u n c t i o n a l form

and parameters f o r a) the p h y s i c o c h e m i c a l ( o r e m p i r i c a l ) model and

b) the e r r o r model r e l a t i n g the e x p e r i m e n t a l o b s e r v a t i o n s t o the
u n d e r l y i n g c h e m i c a l c o m p o s i t i o n . Among the assumptions, o r assumed
parameters t o be t e s t e d , the f o l l o w i n g a r e o f s p e c i a l importance:
Functional Relation
o number o f c h e m i c a l components
o c h a r a c t e r i s t i c s p e c t r a or chromatographic p a t t e r n s
o m a t h e m a t i c a l r e l a t i o n f o r the response f o r each
component ( i n c l u d e s c o r r e c t i d e n t i f i c a t i o n , and
curve shape)
o m a t r i x e f f e c t s and i n t e r f e r e n c e [ i n t e r a c t i o n s ] among
components
o parameters such as the b l a n k , r e c o v e r y , s e n s i t i v i t y
(efficiency)
E r r o r Model
o cumulative d i s t r i b u t i o
o parameters [ v a r i a n c e , h i g h e r moments] ( v a r i a n c e
components f o r compound d i s t r i b u t i o n s )
o a u t o c o r r e l a t i o n [non-white n o i s e ]
o s y s t e m a t i c e r r o r or b i a s [bounds]
o b l u n d e r s ( d i s c r i m i n a t i o n from chance o u t l i e r s , from
discoveries)
H y p o t h e s i s t e s t i n g i s a p p l i c a b l e t o a l l o f the above f a c t o r s .
D e t e c t i o n d e c i s i o n s may be made, f o r example, u s i n g the c r i t i c a l
2
l e v e l o f S t u d e n t ' s - t t o t e s t f o r b i a s , o r the c r i t i c a l l e v e l o f x
to t e s t an assumed s p e c t r a l shape o r c a l i b r a t i o n model o r e r r o r
model. For a g i v e n measurement d e s i g n and assumption t e s t proce-
dure, one can e s t i m a t e the c o r r e s p o n d i n g d e t e c t i o n l i m i t f o r the
a l t e r n a t i v e h y p o t h e s i s , e.g., the minimum d e t e c t a b l e b i a s . As w i t h
a n a l y t e d e t e c t i o n , the a b i l i t y t o d e t e c t erroneous assumptions
r e s t s h e a v i l y on the d e s i g n o f the experiment; and the s t u d y o f
o p t i m a l d e s i g n s i s a f i e l d unto i t s e l f .
A s u r v e y o f s e v e r a l o f the above model-parameter assumptions,
as r e l a t e d t o c h e m i c a l component ( o r a n a l y t e ) d e t e c t i o n w i l l be
presented l a t e r . L e t us t e r m i n a t e t h i s p r e v i e w w i t h two observa-
t i o n s : a) T e s t s o f assumptions may themselves r e s t upon assumptions
-- an obvious case b e i n g the use o f Student's t , w h i c h r e s t s upon
the a s s u m p t i o n o f n o r m a l i t y ; b) D e t e c t i o n o f an a n a l y t e through
2
model f a i l u r e ( l a c k o f f i t ) -- e.g., e v a l u a t i n g x when f i t t i n g a
spectrum w i t h one component m i s s i n g -- i s l e s s s e n s i t i v e than
d i r e c t d e t e c t i o n u s i n g the c o r r e c t model. T h i s i s due t o c o l -
l i n e a r i t y among s p e c t r a l p a t t e r n s ( o r o v e r l a p p i n g chromatographic
peaks) ( 1 3 ) .
1.4 Analyte Detection. T h i s i s a p r i m a r y f o c u s f o r t h i s volume,

the s p e c i f i c a t i o n o f c r i t i c a l l e v e l s o r t h r e s h o l d s f o r a n a l y t e
d e t e c t i o n d e c i s i o n s , and the d e s i g n o f CMP's t o a c h i e v e r e q u i s i t e
analyte detection l i m i t s . The f o l l o w i n g s e c t i o n i n c l u d e s an
h i s t o r i c a l p e r s p e c t i v e on the t o p i c . A t u t o r i a l i s p r o v i d e d i n the
c h a p t e r by K i r c h m e r ( 1 4 ) , where a c r u c i a l d i s t i n c t i o n i s noted:
t h a t i s , the d e t e c t i o n d e c i s i o n i s made i n r e f e r e n c e t o an
observed, random e x p e r i m e n t a l outcome ( e s t i m a t e d c o n c e n t r a t i o n ) ,

whereas the detection limit refers t o the underlying true

c o n c e n t r a t i o n w h i c h t h e CMP i s c a p a b l e o f d e t e c t i n g . The c h i e f
r e a s o n f o r i n t e r e s t i n t h e l a t t e r i s advanced p l a n n i n g and d e s i g n
i . e . , a s s e s s i n g t h e c a p a b i l i t y o f t h e CMP i n q u e s t i o n t o meet
the measurement needs.
Because o f t h e b r o a d scope o f d e t e c t i o n , as o u t l i n e d i n t h e
p r e c e d i n g p a r a g r a p h s , i t i s u s e f u l t o d i s t i n g u i s h some o f t h e
q u a n t i t i e s o r e v e n t s d e t e c t e d w i t h a p p r o p r i a t e symbols. For the
purposes o f t h i s c h a p t e r , t h e f o l l o w i n g w i l l be used:
C r i t i c a l Level Detection Limit
generic symbol Lc
event o r system s t a t e c
(earthquake, o i l s p i l l )
analyte concentratio
( o r amount)
i n s t r u m e n t response Sc
(net s i g n a l )
bias *c
e x t e r n a l random e r r o r
( n o n - P o i s s o n ; "between")
model - l a c k o f f i t --
I n a d d i t i o n t o t h e above, 1^ i s u s e d t o denote t h e e x t e r n a l
l i m i t w h i c h d r i v e s t h e d e s i g n o f t h e Measurement P r o c e s s (MP) .
Thus, i f s u c c e s s f u l p r o c e s s c o n t r o l , o r e a r l y w a r n i n g ( n a t u r a l o r
human d i s a s t e r s ) , o r fundamental c h e m i c a l r e s e a r c h depends on
a c h i e v i n g a l i m i t 1^, t h e n t h e MP must be so d e s i g n e d t h a t i t s
Note t h a t t h e c r i t i c a l l e v e l o f t h e a p p r o p r i a t e t e s t s t a t i s t i c
z t e t
( i-a i-a c ) c a n g e n e r a l l y be used as a n o r m a l i z e d a l t e r n a -
tive t o Xj,, S , e t c .
c The " d e t e c t i o n l i m i t " f o r a t e s t s t a t i s t i c ,
however, i s m e a n i n g l e s s , as x , S , e t c . r e f e r t o t h e t r u e u n d e r l y -
D D
ing quantity. A c o r o l l a r y i s t h a t t h e term " d e t e c t i o n l i m i t " i s

a l s o w i t h o u t meaning i n t h e absence o f an a l t e r n a t i v e h y p o t h e s i s .
( T h i s i s perhaps an o b v i o u s p h i l o s o p h i c a l m a t t e r , b u t i n p r i n c i p l e ,
the n u l l h y p o t h e s i s cannot be r e j e c t e d , e x c e p t by chance [ a - e r r o r ] ,
if no a l t e r n a t i v e e x i s t s ; t h e -error i s t h e n necessarily
undefined. Of c o u r s e an unexpected r e j e c t i o n c a n l e a d t o an
e x c i t i n g search f o r the a l t e r n a t i v e . )
2. HISTORICAL PERSPECTIVE
The d u a l q u e s t i o n s , "How l i t t l e c a n I d e t e c t ? " , and "Has something

been d e t e c t e d ? " have l o n g caught t h e a t t e n t i o n o f a n a l y t i c a l
scientists. Throughout r e c e n t h i s t o r y ( i . e . , 20th c e n t u r y ) a
number o f responses have been f o r m u l a t e d , such as
o The i n t u i t i v e [ f o r m u l a t i o n ] : b a s i n g d e t e c t i o n d e c i s i o n s and
l i m i t s on sound, b u t n o t r e a d i l y q u a n t i f i a b l e e x p e r i e n c e

o The ad hoc: s e l e c t i n g a r i g i d f o r m u l a , o f t e n based on some

reasonable limiting condition, v i a dictum, v o t i n g or
consensus
o The s i g n a l / n o i s e : g e n e r a l l y assuming white noise, and

a d d r e s s i n g p r i m a r i l y t e s t i n g o f an o b s e r v e d s i g n a l
o The a v o i d a n c e : only results o r measurement p r o c e s s e s

t h o r o u g h l y removed from t h e d e t e c t i o n l i m i t deserve o u r
attention
o The h y p o t h e s i s t e s t i n g : where e x p l i c i t a t t e n t i o n i s g i v e n
t o t h e r i s k s o f b o t h f a l s e p o s i t i v e and f a l s e n e g a t i v e
detection decisions.
In reviewing the h i s t o r f detectio limit (i Analytical

Chemistry) i t i s h e l p f u
d i f f e r e n c e s i n mind. I
has meaning, t h e n i t i s e s s e n t i a l t h a t t h e above q u e s t i o n s be f u l l y
d e f i n e d and e x p l i c i t l y addressed. I n the view o f t h i s author a
m e a n i n g f u l approach t o a n a l y t e d e t e c t i o n must be c o n s i s t e n t w i t h
our approach t o u n c e r t a i n t y components o f measurement p r o c e s s e s and
e x p e r i m e n t a l r e s u l t s ; t h e soundest approach i s p r o b a b l y t h e l a s t
[ h y p o t h e s i s t e s t i n g ] tempered w i t h an a p p r o p r i a t e measure o f t h e
first [scientific intuition].
T a b l e I has been p r e p a r e d from t h i s p e r s p e c t i v e . The a u t h o r s
s e l e c t e d a r e drawn p r i m a r i l y from those who have c o n t r i b u t e d b a s i c
s t a t e m e n t s on t h e i s s u e o f d e t e c t i o n c a p a b i l i t i e s o f c h e m i c a l
measurement p r o c e s s e s [ " d e t e c t i o n l i m i t s " ] , as opposed t o s i m p l y
addressing d e t e c t i o n d e c i s i o n s f o r observed r e s u l t s ["critical
levels"]. I n f a i r n e s s t o those n o t l i s t e d , i t i s i m p o r t a n t t o note
t h a t a) a s e l e c t i o n o n l y , spanning t h e l a s t s e v e r a l decades has
been g i v e n , and t h a t b) t h e r e a l s o e x i s t many e x c e l l e n t a r t i c l e s
(15.16^ and books (12.17.18^ which r e v i e w t h e t o p i c . It is
i m m e d i a t e l y c l e a r from T a b l e I t h a t t h e t e r m i n o l o g y has been wide
r a n g i n g , even i n those cases where t h e c o n c e p t u a l b a s i s ( h y p o t h e s i s
t e s t i n g ) has been i d e n t i c a l . Nomenclature, u n l i k e s c i e n t i f i c f a c t s
and c o n c e p t s , c a n be approached, however, t h r o u g h consensus. The
I n t e r n a t i o n a l U n i o n o f Pure and A p p l i e d C h e m i s t r y [IUPAC], which
appears t w i c e i n T a b l e I , i s t h e i n t e r n a t i o n a l body o f c h e m i s t s
charged w i t h t h i s r e s p o n s i b i l i t y . A t t h i s p o i n t i t w i l l be h e l p f u l
t o examine t h e p o s i t i o n o f IUPAC as w e l l as t h e c o n t r i b u t i o n s o f
some o f t h e o t h e r a u t h o r s c i t e d i n T a b l e I .
F r i t z F e i g l ( 1 9 ) , the f a t h e r o f "Spot T e s t s , " heads t h e l i s t
p r i m a r i l y as one who suggested lower l i m i t s f o r c h e m i c a l measure-
ment, h e r e t r a n s l a t e d (from t h e german) as " i d e n t i f i c a t i o n l i m i t s , "
which represented the best experience [or chemical i n t u i t i o n ] o f
the day. Such l i m i t s , t y p i c a l l y i n t h e microgram range, were
s c a r c e l y ad hoc, b u t they o f c o u r s e l a c k e d t h e s t a t i s t i c a l s o p h i s -
t i c a t i o n o f l a t t e r day l i m i t s . F e i g l ' s l i m i t s , however, deserve
our a t t e n t i o n even today, i n t h a t t h e y r e c o g n i z e t h e o v e r a l l
c a p a b i l i t y o f t h e measurement p r o c e s s i n c l u d i n g t h a t which cannot
be r e a d i l y t r e a t e d by s t a t i s t i c s [Note__2] .

Table I . H i s t o r i c a l P e r s p e c t i v e -- D e t e c t i o n L i m i t Terminology
Feigl ('23) I d e n t i f i c a t i o n L i m i t (19)

Altshuler ('63) Minimum D e t e c t a b l e True A c t i v i t y (21)
K a i s e r ('65-'68) - L i m i t o f Guarantee f o r P u r i t y (20)
S t . J o h n ('67) L i m i t i n g Detectable Concentration ( S / N r m s )
(24)
C u r r i e ('68) D e t e c t i o n L i m i t (23)
N i c h o l s o n ('68) D e t e c t a b i l i t y (25)
IUPAC ('76) L i m i t o f D e t e c t i o n (29)
I n g l e ('74) ("[too] complex...not common") (27)
Lochamy ('76) Minimum D e t e c t a b l e A c t i v i t y (92)
G r i n z a i d ('77) Nonparametric D e t e c t i o n L i m i t (26)
L i t e a n u ('80) F r e q u e n t o m e t r i c D e t e c t i o n L i m i t (8)
NRC ('84) Lower L i m i t o f D e t e c t i o n [28]
IUPAC ('86) D e t e c t i o n L i m i t (30)
IAEA ('87)
Among the o t h e r s c i t e d i n Table I , K a i s e r (20) deserves major

c r e d i t f o r i n t r o d u c i n g the h y p o t h e s i s t e s t i n g concept i n t o s p e c t r o -
c h e m i c a l a n a l y s i s , as does A l t s h u l e r (21.) i n r a d i o a c t i v i t y measure-
ment. W i l s o n (22.) championed i t s use f o r w a t e r a n a l y s i s , and
C u r r i e (.23) p r o v i d e d an approach f o r d e t e c t i o n and q u a n t i f i c a t i o n
i n a n a l y t i c a l and r a d i o c h e m i s t r y . The r e f e r e n c e by S t . J o h n (24)
has been one o f the most c i t e d o f those based on s i g n a l / n o i s e ,
though i t does n o t address the e r r o r o f the second k i n d ( f a l s e
negative). N i c h o l s o n (25) gave one o f the e a r l i e s t t r e a t m e n t s f o r
extreme l o w - l e v e l ( P o i s s o n ) c o u n t i n g d a t a , and G r i n z a i d (26)
o f f e r e d a r o b u s t t r e a t m e n t n o t r e q u i r i n g the assumption o f any
specific distribution. L i t e a n u ' s f r e q u e n t o m e t r i c method (8) was
a l s o d i s t r i b u t i o n - f r e e , i n the sense t h a t an e x p e r i m e n t a l e s t i m a t e
o f the d e t e c t i o n l i m i t was d e r i v e d from the observed f r a c t i o n of
f a l s e negatives, using a r e g r e s s i o n technique. The paper by I n g l e
( 2 7 ) , w h i c h was o b v i o u s l y d e s i g n e d t o be t u t o r i a l ( p u b l i s h e d i n the
J . Chemical E d u c a t i o n ) i s noteworthy i n t h a t i t suggested t h a t the
concept o f the e r r o r o f the second k i n d (which i s i n t r i n s i c t o the
s t a t i s t i c a l t h e o r y o f h y p o t h e s i s t e s t i n g ) was s i m p l y too complex
f o r o r d i n a r y c h e m i s t s t o grasp! R e g r e t t a b l y , t h e r e seems t o be
some s u p p o r t f o r such a statement; b u t H y p o t h e s i s T e s t i n g i s one o f
the k e y s t o n e s o f e v e r y elementary c o u r s e i n S t a t i s t i c s , so i t s
f o r m a l i n t r o d u c t i o n i n t o the e d u c a t i o n o f the a n a l y t i c a l chemist
would seem not too e s o t e r i c a s t e p .
An e x h a u s t i v e r e v i e w o f the d e f i n i t i o n and a p p l i c a t i o n o f
D e t e c t i o n L i m i t s f o r n u c l e a r and a n a l y t i c a l c h e m i c a l measurements
was p u b l i s h e d i n 1984 (28) . The r e a d e r may w i s h t o scan the t i t l e s
o f the papers t h e r e c i t e d , t o g a i n f u r t h e r i n s i g h t r e g a r d i n g b a s i c
p r i n c i p l e s and t e r m i n o l o g y , c o u n t i n g s t a t i s t i c s , n o n - c o u n t i n g and
non-normal random e r r o r s , random and s y s t e m a t i c v a r i a t i o n s i n the
b l a n k , B a y e s i a n approaches, r e p o r t i n g , a v e r a g i n g and c e n s o r i n g
treatments, o p t i m i z a t i o n , i n f l u e n c e of alternative spectrum
deconvolution techniques, etc. I n the body o f Ref. 28 s p e c i a l
a t t e n t i o n i s g i v e n a l s o t o t o p i c s such as s i m p l e and multicomponent
n u c l e a r spectrum f i t t i n g and extreme l o w - l e v e l c o u n t i n g .
W i t h r e s p e c t t o IUPAC, b o t h the p o s i t i o n p u b l i s h e d i n 1976
( 2 9 ) , w h i c h a d d r e s s e d nomenclature, symbols and u n i t s i n a n a l y t i c a l

o p t i c a l s p e c t r o s c o p y , and the more g e n e r a l a n a l y t i c a l nomenclature

document, now i n r e v i e w ( 1 0 ) , t r e a t d e t e c t i o n from the h y p o t h e s i s
t e s t i n g v i e w p o i n t . A n o n - c o n c e p t u a l d i f f e r e n c e l i e s i n the c h o i c e
o f the r i s k l e v e l ( f a l s e p o s i t i v e s and n e g a t i v e s ) . The 1976
r e p o r t , w h i c h grew out o f K a i s e r ' s work, used a f i x e d v a l u e o f 3.00
f o r the s t a n d a r d d e v i a t i o n m u l t i p l i e r (S/N) f o r d e t e c t i o n d e c i -
s i o n s . T h i s would c o r r e s p o n d t o a f a l s e p o s i t i v e r i s k o f 1-0.9986,
o r 0.14% ( 1 - s i d e d t e s t ) , i f the p o p u l a t i o n were normal, and a
known. (The f a l s e n e g a t i v e r i s k p was n o t e x p l i c i t l y t r e a t e d . )
The c u r r e n t IUPAC Nomenclature Document recommends r i s k l e v e l s ( a ,
P) o f 5%, c o r r e s p o n d i n g t o a m u l t i p l i e r o f 1.645 f o r a known,
normal p o p u l a t i o n . Both documents r e c o g n i z e the e f f e c t s o f v a r y i n g
degrees o f freedom i n e s t i m a t i n g the v a r i a n c e o f the b l a n k ; the
l a t t e r document s p e c i f i c a l l y recommends the use o f S t u d e n t ' s - t t o
compensate, j u s t as i s done i n the c o n s t r u c t i o n o f normal
confidence i n t e r v a l s .
The h i s t o r i c a l e v o l u t i o
v e r y u n f o r t u n a t e and n e e d l e s
concept. Awareness o f the n a t u r e o f t h i s c o n f u s i o n i s c r u c i a l , i f
we as a n a l y t i c a l s c i e n t i s t s are t o a r r i v e a t a common and meaning-
f u l approach t o d e t e c t i o n , an approach t h a t can s e r v e s o c i e t y
r a t h e r than add an e x t r a l e v e l o f c o n f u s i o n t o a t o p i c w h i c h the
p u b l i c r e g a r d s as a l r e a d y c o m p l i c a t e d , a l b e i t i m p o r t a n t .
The f a c t s are t h a t f o r a t l e a s t the l a s t decade o r two t h e r e
has been b r o a d i n t e r n a t i o n a l s u p p o r t f o r the h y p o t h e s i s t e s t i n g
framework f o r making a n a l y t e d e t e c t i o n d e c i s i o n s , and e v a l u a t i n g --
e s p e c i a l l y f o r purposes o f d e s i g n and p l a n n i n g -- the i n h e r e n t
d e t e c t i o n c a p a b i l i t i e s o f measurement p r o c e s s e s . In t h i s context,
a number o f a u t h o r s and i n s t i t u t i o n s have employed terms l i k e
" d e t e c t i o n l i m i t " ( o r " l i m i t o f d e t e c t i o n " ) t o denote the l a t t e r ,
inherent detection c a p a b i l i t y , generally i n u n i t s of concentration
o r amount (8.12.17. 21-23. 25.30.36). I n the e a r l i e s t work o f some
who most s t r o n g l y came t o s u p p o r t the h y p o t h e s i s t e s t i n g model,
however, the n o t i o n o f the f a l s e n e g a t i v e [0 - e r r o r ] d i d n o t
appear ( 3 1 ) . K a i s e r i n p a r t i c u l a r l a b e l e d h i s t h r e s h o l d l e v e l "Die
Nachweisgrenze," o r D e t e c t i o n L i m i t . I n 1965 K a i s e r t r e a t e d the
second k i n d o f e r r o r (p) , and i n t r o d u c e d "Die G a r a n t i e g r e n z e f u r
R e i n h e i t " as the c o r r e s p o n d i n g t r u e c o n c e n t r a t i o n l e v e l f o r the
a l t e r n a t i v e hypothesis (32). K a i s e r ' s impact on the f i e l d o f
A n a l y t i c a l C h e m i s t r y has been e x t r e m e l y s i g n i f i c a n t , and i t i s not
s u r p r i s i n g t h a t many chemists have adopted h i s t e r m i n o l o g y f o r the
Detection Limit. I t has been adopted, however, i n many cases t o
i n d i c a t e not j u s t the s i g n a l / n o i s e l e v e l f o r making d e t e c t i o n
d e c i s i o n s , b u t a l s o as a measure o f the i n h e r e n t d e t e c t i o n
c a p a b i l i t y o f the measurement p r o c e s s i n q u e s t i o n . S i n c e the e r r o r
o f the second k i n d [p] e x i s t s whether i t ' s r e c o g n i z e d o r n o t , t h i s
p r a c t i c e has l e d t o a de f a c t o f a l s e n e g a t i v e r i s k o f 50% -- a
v a l u e w h i c h i s t o t a l l y out o f b a l a n c e w i t h a f a l s e p o s i t i v e r i s k o f
0.14%, o r even 5%! T h o u g h t f u l and l u c i d c r i t i q u e o f t h i s m a t t e r
may be found i n R e f ' s . 8 (p. 263) and 14. A c u r i o u s footnote to
t h i s d i s c u s s i o n i s t h a t one s c a r c e l y ever encounters K a i s e r ' s
second term, " L i m i t o f Guarantee f o r P u r i t y , " i n the s c i e n t i f i c
literature.

On the s u b j e c t o f nomenclature, a word c o n c e r n i n g h i s t o r i c a l l y

u s e d terms f o r the d e t e c t i o n decision point or l e v e l i s i n o r d e r .
As s t a t e d i m m e d i a t e l y above, a number o f a n a l y s t s , f o l l o w i n g
K a i s e r , use " L i m i t o f D e t e c t i o n " or " D e t e c t i o n L i m i t " as both the
measure o f ( t r u e c o n c e n t r a t i o n ) d e t e c t i o n c a p a b i l i t y and as a
s t a t i s t i c a l c r i t i c a l l e v e l o r t h r e s h o l d t o make d e t e c t i o n d e c i -
sions. F o l l o w i n g e s t a b l i s h e d p r a c t i c e i n S t a t i s t i c s , the term
" C r i t i c a l L e v e l " was recommended i n ( 2 3 ) . " C r i t e r i o n of Detection"
has been employed by W i l s o n ( 2 2 ) ; and L i t e a n u ( 8 ) , who speaks o f
the " d e c i s i o n c r i t e r i o n " as a s t r a t e g y , terms the numerical
c o m p a r i s o n l e v e l the " D e c i s i o n ( o r D e t e c t i o n ) Threshold."
The g r e a t m a j o r i t y o f the a u t h o r s c i t e d i n the foregoing
d i s c u s s i o n emphasized t h a t the d e t e c t i o n l i m i t must r e f e r t o the
e n t i r e a n a l y t i c a l measurement p r o c e s s . I n many cases one f i n d s
t h a t not the case -- i . e . , workers may r e f e r (sometimes appro-
p r i a t e l y and i n t e n t i o n a l l y ) t o j u s t the i n s t r u m e n t a l measurement
s t e p , or t o i d e a l , pur
may be f a r too o p t i m i s t i
t i o n s i n d i c a t e " t y p i c a l " d e t e c t i o n l i m i t s , an a c c e p t a b l e p r a c t i c e
provided the measurement p r o c e s s and sample n a t u r e ( i n c l u d i n g
matrix and i n t e r f e r e n c e e f f e c t s ) are rigidly controlled and
s u b j e c t e d t o a p p r o p r i a t e ruggedness t e s t i n g .
2.1 P r e s e n t S t a t e o f the A r t . A p e r u s a l o f the analytical

l i t e r a t u r e two decades ago r e v e a l e d c o n s i d e r a b l e d i s p a r i t y i n the
s p e c i f i c a t i o n of detection l i m i t s . T h i s i s shown i n F i g . 3 w h i c h
i s r e p r o d u c e d from ( 2 3 ) . Then c u r r e n t d e f i n i t i o n s spanned n e a r l y 3
o r d e r s o f magnitude when a p p l i e d t o the same measurement problem!
Concern f o r such d e f i n i t i o n a l (and/or c o n c e p t u a l ) d i s p a r i t y has l e d
a number o f n a t i o n a l and i n t e r n a t i o n a l o r g a n i z a t i o n s t o address the
need f o r a common, r a t i o n a l b a s i s f o r t r e a t i n g t h i s m a t t e r .
Because o f c o n c e n t r a t i o n r e l a t e d e f f e c t s o f t r a c e c h e m i c a l s p e c i e s
on h e a l t h , p r o p e r t i e s o f h i g h p u r i t y m a t e r i a l s , and even g l o b a l
climate, relative detection limits f o r d i f f e r e n t measurement
p r o c e s s e s are not enough; d e t e c t i o n c a p a b i l i t i e s must be a s s e s s e d
i n a b s o l u t e u n i t s . Awareness o f the c o n f u s i o n s u r r o u n d i n g detec-
t i o n l i m i t p r a c t i c e s , by o r g a n i z a t i o n s such as IUPAC, IAEA, ACS, a
number o f US r e g u l a t o r y a g e n c i e s , and more r e c e n t l y CODATA (Commit-
tee on Data f o r S c i e n c e and Technology) i s a v e r y p o s i t i v e t h i n g .
The d i f f i c u l t y and importance o f the t a s k i s h i g h l i g h t e d by s e v e r a l
o f the a u t h o r s i n t h i s volume, n o t a b l y : a) Crummett (33) ["In
s p i t e o f e x t r a o r d i n a r y e f f o r t s (on the p a r t o f s c i e n t i f i c s o c i e t i e s
to p r o p e r l y define d e t e c t i o n l i m i t s ) a n a l y s t s continue to present
t h e i r r e s u l t s i n forms w h i c h cause the c r e d i b i l i t y o f the d a t a to
be q u e s t i o n e d o r the meaning t o be m i s i n t e r p r e t e d " ] ; b) Brossman
(34) ["Attempts by our t a s k f o r c e on l o w - l e v e l d a t a t o make a
rigorous conceptual and s t a t i s t i c a l comparison ... have been
unsuccessful. Even s i m i l a r terms are d e f i n e d in different,
non-comparable ways..."]; and c) C u r r i e and P a r r ( 3 5 ) , where i t was
o b s e r v e d t h a t i n t e r n a t i o n a l i n t e r l a b o r a t o r y comparisons i n v o l v i n g
the same b i o e n v i r o n m e n t a l reference materials r e s u l t e d i n mutually
e x c l u s i v e r e s u l t s . For example, q u a n t i t a t i v e r e s u l t s f o r a r s e n i c
( i n h o r s e k i d n e y ) were r e p o r t e d by some l a b o r a t o r i e s a t l e v e l s
w h i c h exceeded the " d e t e c t i o n l i m i t s " o f o t h e r l a b o r a t o r i e s (which
detected no a r s e n i c ) by as much as 4 o r d e r s o f magnitude!

5000 I 1 1 1 1 1 1
DEFINITIONS:
1 BACKGROUND STANDARD DEVIATION (o ) B
2000 2 - 10% OF THE BACKGROUND
3 - 2o B
1000 4- 3og
oq
o 500
a
8 - 100 dps
200
z
100
o
o -(LQ)-Z
UJ
I- 50
UJ
a
20 -(LD)-
10 -(L )-
C
O
5 _1_
3 4 5 6 7
DEFINITION
F i g . 3. Ordered d e t e c t i o n l i m i t s 1968. (Reproduced from

Ref. 23. C o p y r i g h t 1968 American Chemical S o c i e t y . )

I n c i d e n t a l l y , the Brossman t a s k f o r c e e p i t o m i z e s a v e r y s e r i o u s
problem: the c o d i n g o f d a t a i n t o c o m p u t e r i z e d d a t a bases. Such
d a t a bases w i l l d o u b t l e s s have s i g n i f i c a n t d i s t r i b u t i o n s and
l i f e t i m e s , so the e f f e c t s o f p o s s i b l e d i s t o r t i o n s and i n f o r m a t i o n
l o s s w i l l be u n f o r t u n a t e l y a m p l i f i e d .
Understanding and acceptance of the h y p o t h e s i s - t e s t i n g
p o s i t i o n t a k e n by IUPAC (29.30). the US N u c l e a r R e g u l a t o r y Commis-
s i o n [ 2 8 ] , the UK Water Research Centre ( 3 6 ) , the IAEA, and
r e f l e c t e d i n many o f the r e c e n t t e x t s i n A n a l y t i c a l C h e m i s t r y and
Chemometrics ( 3 7 ) , promises t o r e s o l v e the n e e d l e s s , c u r r e n t
disarray. Some o f the c u r r e n t d i v e r s i t y can be seen i n F i g . 4,
which p r e s e n t s f o u r o f the p r i n c i p a l d e t e c t i o n l i m i t d e f i n i t i o n s i n
vogue (and/or i n r e g u l a t o r y g u i d e s ) i n the U.S. Comparisons among
the s t a t e m e n t s , t o g e t h e r w i t h the s u p p o r t i n g documents, show t h a t :
(1) the p e r r o r ( f a l s e n e g a t i v e ) i s i g n o r e d i n a l l b u t one, c a u s i n g
i t t o assume a de f a c t o v a l u e 50%; (2) t r e a t m e n t o f the b l a n k i s
ambiguous o r absent i
the r e a g e n t b l a n k i n a
e r r o r , t a k i n g i n t o account the number o f degrees o f freedom and
Student's t i s l a c k i n g . There i s some i r o n y i n the f a c t t h a t the
f o u r t h d e f i n i t i o n s t a t e s t h a t the LOD i s "the l o w e s t c o n c e n t r a t i o n
... s t a t i s t i c a l l y d i f f e r e n t from a b l a n k " , i n v i e w o f a comment i n
the r e f e r e n c e c i t e d (Long and W i n e f o r d n e r , 1983). These a u t h o r s
note t h a t the " w e l l - b a s e d b u t seldom used concept i n the c a l c u l a -
t i o n o f d e t e c t i o n l i m i t s ... the l i m i t o f guarantee f o r p u r i t y , c , G
d e s c r i b e d by K a i s e r ... [ r e p r e s e n t s ] the l o w e s t statistically

discernable signal." Long and W i n e f o r d n e r go on t o show t h a t the
o r i g i n a l IUPAC d e f i n i t i o n (29) and the LOD o f r e f e r e n c e (16) i n d e e d
y i e l d 50% f a l s e n e g a t i v e s (P) . Kaiser's c , incidentally, i s
G
c o n c e p t u a l l y i d e n t i c a l t o C u r r i e ' s 1^ (23) and Boumans' " L i m i t o f

I d e n t i f i c a t i o n " (15).
I n a d d i t i o n t o the above, one c o n t i n u a l l y f i n d s v a r y i n g ad hoc
o r even u n d e f i n e d usage i n the p e e r - r e v i e w e d a n a l y t i c a l l i t e r a t u r e .
For example, t h r e e r e c e n t p a p e r s , examined because o f i n t e r e s t i n
t h e i r c h e m i c a l c o n t e n t , a l l deemed d e t e c t i o n l i m i t s o f s u f f i c i e n t
importance t o i n c l u d e n u m e r i c a l t a b u l a t i o n s . However, the f i r s t
a u t h o r s t a t e d t h a t h i s d e t e c t i o n l i m i t r e p r e s e n t e d a s i g n a l to
n o i s e r a t i o o f 10; the second d e f i n e d i t as t w i c e the s t a n d a r d
d e v i a t i o n o f the background s i g n a l ; and the t h i r d gave no
i n d i c a t i o n as t o h i s meaning.
I n c o n c l u s i o n , i t i s u r g e n t t h a t the a n a l y t i c a l community
adopt a u n i f o r m and d e f e n s i b l e approach t o the concept o f detec-
tion. A p a r t from ad hoc or u n s t a t e d p r o c e d u r e s , f a i l u r e t o
r e c o g n i z e the e r r o r o f the second k i n d [p] -- i . e . , f a i l u r e t o
d i s t i n g u i s h between d e t e c t i o n d e c i s i o n s and d e t e c t i o n c a p a b i l i t i e s
-- i s the most s e r i o u s c o n c e p t u a l f a u l t , p l a c i n g f a l s e n e g a t i v e s a t
the l e v e l o f c o i n - f l i p p i n g a c c u r a c y . F a i l u r e t o t a k e i n t o account
a l l major s o u r c e s o f e r r o r , e s p e c i a l l y the n a t u r e o f the b l a n k , i s
the most s e r i o u s measurement f a u l t . A r e v i e w o f some o f the more
critical assumptions and technical issues related to valid
detection l i m i t s follows.

Lower L i m i t o f D e t e c t i o n ( L L P ) . "The LLD i s d e f i n e d , f o r purposes

o f these s p e c i f i c a t i o n s , as the s m a l l e s t c o n c e n t r a t i o n o f the
r a d i o a c t i v e m a t e r i a l i n a sample t h a t w i l l y i e l d a n e t count,
above system background, t h a t w i l l be d e t e c t e d w i t h 95%
p r o b a b i l i t y w i t h o n l y 5% p r o b a b i l i t y o f f a l s e l y c o n c l u d i n g
that a blank observation represents a " r e a l " s i g n a l " (94).
I n s t r u m e n t a l D e t e c t i o n L i m i t ( I D L ) . "The c o n c e n t r a t i o n e q u i v a l e n t
t o a s i g n a l , due t o the a n a l y t e , w h i c h i s e q u a l t o t h r e e times
the s t a n d a r d d e v i a t i o
ments o f a r e a g e n t
Method D e t e c t i o n L i m i t (MDL^. "The method d e t e c t i o n l i m i t (MDL) i s

d e f i n e d as the minimum c o n c e n t r a t i o n o f a s u b s t a n c e t h a t can
be measured and r e p o r t e d w i t h 99% c o n f i d e n c e t h a t the a n a l y t e
c o n c e n t r a t i o n i s g r e a t e r t h a n z e r o and i s d e t e r m i n e d from
a n a l y s i s o f a sample i n a g i v e n m a t r i x c o n t a i n i n g the a n a l y t e "
(%).
L i m i t o f D e t e c t i o n CLOD'). "The l i m i t o f d e t e c t i o n (LOD) i s d e f i n e d

as the l o w e s t c o n c e n t r a t i o n l e v e l t h a t can be d e t e r m i n e d t o be
s t a t i s t i c a l l y d i f f e r e n t from a b l a n k . The c o n c e p t i s r e v i e w e d
i n [ r e f . 38] t o g e t h e r w i t h the s t a t i s t i c a l b a s i s f o r i t s
e v a l u a t i o n . A d d i t i o n a l c o n c e p t s i n c l u d e method d e t e c t i o n
l i m i t (MDL), w h i c h r e f e r s t o the l o w e s t c o n c e n t r a t i o n o f
a n a l y t e t h a t a method can d e t e c t r e l i a b l y i n e i t h e r a sample
o r b l a n k , and the i n s t r u m e n t d e t e c t i o n l i m i t ( I D L ) , w h i c h
r e f e r s t o the s m a l l e s t s i g n a l above background n o i s e t h a t an
i n s t r u m e n t can d e t e c t r e l i a b l y . Sometimes, the IDL and LOD
are o p e r a t i o n a l l y the same. I n p r a c t i c e , an i n d i c a t i o n o f
whether an a n a l y t e i s d e t e c t e d by an i n s t r u m e n t i s sometimes
b a s e d on the e x t e n t o f w h i c h the a n a l y t e s i g n a l exceeds peak-
to-peak n o i s e " (16).
F i g . 4. Four D e f i n i t i o n s f o r D e t e c t i o n Limits Related to

C u r r e n t U.S. R e g u l a t o r y P r a c t i c e .

3. DEPTH: L i m i t a t i o n s , Assumptions, and T e c h n i c a l I s s u e s
The f o r e g o i n g t e x t r e p r e s e n t s a b r i e f o v e r v i e w o f some o f t h e
s o c i e t a l , h i s t o r i c a l , and b r o a d c o n c e p t u a l i s s u e s r e l a t i n g t o
d e t e c t i o n and c h e m i c a l measurements. Here we o f f e r an o v e r v i e w , i n
c a t a l o g o r d i c t i o n a r y format, o f a s e r i e s o f t e c h n i c a l i s s u e s
d i r e c t l y r e l a t e d t o t h e e s t i m a t i o n and v a l i d i t y o f a n a l y t e detec-
tion limits. B a l a n c e d coverage has been t h e i n t e n t , b u t s p e c i a l
a t t e n t i o n has been g i v e n t o t o p i c s n o t c o v e r e d elsewhere i n t h i s
volume, and t o q u e s t i o n s a r i s i n g i n d i s c u s s i o n s o r p u t by " u s e r s "
of detection l i m i t s . I n some c a s e s , t h i s l e d t o t h e i n t r o d u c t i o n
o f new m a t e r i a l such as m u l t i p l e d e c i s i o n s and p r o b a b i l i s t i c
pattern detection, u t i l i z a t i o n o f p h y s i c a l c o n s t r a i n t s (on
v a r i a n c e ) , and some e f f e c t s o f v a r y i n g p r o b a b i l i t y d e n s i t y func-
t i o n s [pdf] as r e l a t e d t o e x p e r i m e n t a l d e s i g n and a v a r i a t i o n . The
d i s c u s s i o n i s d i v i d e d i n t o three p a r t s : the f i r s t considering
issues a f f e c t i n g the v a l i d i t
s i s t e s t i n g ] ; t h e second
d e t e c t i o n f o r the a l t e r n a t i v e hypothesis; the t h i r d , c o n s i d e r i n g
m u l t i p l e d e t e c t i o n d e c i s i o n s and D i s c r i m i n a t i o n L i m i t s f o r c h e m i c a l
s p e c i e s and c h e m i c a l p a t t e r n s . A guide t o t h e t o p i c s p r e s e n t e d i n
t h i s s e c t i o n i s g i v e n i n F i g . 5 [Note_3].
3.1 N u l l H y p o t h e s i s T e s t i n g -- Assumptions and C o n c l u s i o n s . When

an experiment i s performed, we t e s t t h e e x p e r i m e n t a l r e s u l t ( x ) by
c o m p a r i s o n t o t h e c r i t i c a l l e v e l o r t h r e s h o l d (x^,) t o d e c i d e
whether o r n o t a n a l y t e has been d e t e c t e d i n excess o f t h e b l a n k o r
background l e v e l . Q u i t e a p a r t from q u e s t i o n s i n v o l v i n g t h e
a l t e r n a t i v e h y p o t h e s i s o r d e t e c t i o n l i m i t , two c r u c i a l p o i n t s must
be k e p t i n mind c o n c e r n i n g t h e n a t u r e and v a l i d i t y o f such a t e s t .
The f i r s t i s t h a t the S t a t i s t i c a l Test f o r S i g n i f i c a n c e , a t
s i g n i f i c a n c e l e v e l a, i s b a s e d on e x a c t l y t h e same p r i n c i p l e s as
the more f a s h i o n a b l e c a l c u l a t i o n o f Confidence Intervals, at
c o n f i d e n c e l e v e l 1-a ( 3 9 ) . ( I n b o t h c a s e s , o f c o u r s e , one must pay
a t t e n t i o n t o 1- v s 2 - s i d e d t e s t s o r i n t e r v a l s . ) The second p o i n t
f o l l o w s : t h a t assumptions a f f e c t i n g t h e v a l i d i t y o f e x p e r i m e n t a l
c o n f i d e n c e i n t e r v a l s a r e j u s t as i m p o r t a n t f o r t h e v a l i d i t y o f
s i g n i f i c a n c e t e s t s . Assumptions w h i c h demand o u r a t t e n t i o n i n c l u d e
the following: control [i.e., e x i s t e n c e ] o f t h e measurement
p r o c e s s ; p o s s i b l e s y s t e m a t i c model [ f u n c t i o n a l ] o r measurement
e r r o r ; and p r o p e r t i e s o f t h e random component ( o r components) o f
e r r o r -- i . e . , form o f t h e d i s t r i b u t i o n [pdf o r c d f ] , parameters o f
2
the d i s t r i b u t i o n [ a , . . . ] , " c o l o r " o f t h e n o i s e ( o r n o i s e power
s p e c t r u m ) , and n o n - s t a t i o n a r i t y (changes w i t h time) and h e t e r o -
s c e d a s t i c i t y (changes w i t h c o n c e n t r a t i o n o r o t h e r experimental
p a r a m e t e r s ) . E r r o r s may a r i s e a l s o from uncompensated changes i n
the measurement p r o c e s s i t s e l f , such as a l t e r a t i o n o f t h e c a l i b r a -
t i o n [ f u n c t i o n a l ] model o r random e r r o r model [ c d f ] due t o c h e m i c a l
m a t r i x e f f e c t s o r i n t e r f e r e n c e . No l e s s i m p o r t a n t a r e t h e computa-
t i o n a l and d a t a r e p o r t i n g s t r a t e g i e s ; these r e p r e s e n t i n t r i n s i c
p a r t s o f t h e o v e r a l l measurement p r o c e s s . A brief catalog of
s e l e c t e d h y p o t h e s i s t e s t i n g and e x p e r i m e n t a l r e s u l t r e l a t e d i s s u e s
follows.

CURRIE Overview of Historical, Societal, and Technical Issues
D e t e c t i o n D e c i s i o n s , a - e r r o r T3.11
(assumptions, v a l i d i t y )
-BASIC ERROR ISSUES

s y s t e m a t i c e r r o r [3.1.1]
normal random e r r o r [3.1.2-.3]
non-normal [3.1.4-.5]
p a i r e d comparisons [3.1.6]
-MEASUREMENT PROCESS ISSUES

background, b a s e l i n e , b l a n k [3.1.7]
e r r o r components, t r u n c a t i o n [3.1.8-.10]
e v a l u a t i o n p r o c e s s , c a l i b r a t i o n [3.1.11-.12]
r e p o r t i n l o w - l e v e l dat [3.1.13]
artificia
A n a l y t e D e t e c t i o n L i m i t , p-error T3.21
( e s t i m a t i o n , power)
-DETECTION LIMITS AND POWER

i g n o r a n c e o f p-error [3.2.1]
l o w e r , upper Lp's [3.2.2]
a, p c o n n e c t i o n (ROC) [3.2.3]
power o f t h e t - t e s t [3.2.4]
-UNCERTAINTY I N Lp [3.2.5-.6]
-SPECIAL TOPICS
o p t i m i z a t i o n [3.2.7]
multicomponent d e t e c t i o n [3.2.8]
random e r r o r v a r i a t i o n [3.2.9]
q u a l i t y ( a l g o r i t h m s , c o n t r o l s ) [3.2.10-.11]
s c r i m i n a t i o n L i m i t : M u l t i p l e D e c i s i o n s T3.31
-DISCRIMINATION LIMITS
l o w e r and upper r e g u l a t o r y l i m i t s [3.3.1]
i m p u r i t y d e t e c t i o n [3.3.2]
MULTIPLE DECISIONS, IDENTIFICATION

m u l t i p l e d e t e c t i o n d e c i s i o n s [3.3.3]
m u l t i c h a n n e l i d e n t i f i c a t i o n [3.3.4]
m u l t i v a r i a b l e p a t t e r n s [3.3.5-.6]
F i g . 5. T o p i c a l Guide t o T e c h n i c a l Overview.

3.1.1 s y s t e m a t i c and model e r r o r . Bounds f o r uncompensated,

non-random e r r o r s must be a l l o w e d f o r by a c o r r e s p o n d i n g i n c r e a s e
i n the c r i t i c a l l e v e l o r c o n f i d e n c e i n t e r v a l . T h i s makes a an
upper l i m i t i f the s y s t e m a t i c bounds, w h i c h need n o t be symmetric,
are g i v e n as upper l i m i t s . The v a l i d i t y o f the c o r r e s p o n d i n g
u n c e r t a i n t y i n t e r v a l c l e a r l y depends h e a v i l y on the Chemical
Intuition or scientific e x p e r t i s e employed, f o r example i n
i d e n t i f y i n g the range o f p o s s i b l e a l t e r n a t i v e models.
a
3.1.2 normal ( w h i t e ) random n o i s e . I f a i s known, l c i - a o , = z
where o Q r e p r e s e n t s the s t a n d a r d d e v i a t i o n o f the e s t i m a t e d n e t

s i g n a l (21). I f " s i m p l e " d e t e c t i o n [gross s i g n a l - b l a n k ] i s
i n v o l v e d , where the b l a n k i s e s t i m a t e d from n e q u i v a l e n t observa-
t i o n s , and the g r o s s s i g n a l from one, then o = a 7 ( n + l ) / n - - CT
Q B B
b e i n g the s t a n d a r d d e v i a t i o n o f the b l a n k . [Note t h a t z e r o

adjustment, as f o r the n u l l l e v e l o r b a s e l i n e o f a ( r e c o r d i n g )
galvanometer, chromatograph spectrophotometer etc. does not
e l i m i n a t e the need f o
propagation. Of c o u r s e
l a r g e span o f l i n e a r b a s e l i n e i s q u i t e p r e c i s e l y a d j u s t e d , f o r
example, by l e a s t squares f i t t i n g o r by g r a p h i c o r even " e y e b a l l "
subtraction. A related point: the ( d e t e c t i o n ) t e s t must be
a p p l i e d t o the n e t s i g n a l , or e q u i v a l e n t c o n c e n t r a t i o n e s t i m a t e ,
because o n l y t h a t has an e x p e c t e d v a l u e under the n u l l h y p o t h e s i s
of e x a c t l y zero. I m p r e c i s e knowledge o f the mean v a l u e f o r the
b l a n k d i s t r i b u t i o n p r e v e n t s a r i g o r o u s t e s t b e i n g a p p l i e d t o the
gross s i g n a l . See comments below on the background, b a s e l i n e , and
blank.]
3.1.3 o unknown (normal) . Student's t r e p l a c e s z when a i s
e s t i m a t e d by r e p l i c a t i o n . 1^, now e q u a l s t s . [Note t h a t 1^, h e r e ,
u n l i k e Lp, i s no l o n g e r a c o n s t a n t , because s ( e s t i m a t e o f o) i s a
random v a r i a b l e . Note a l s o t h a t o ( o r s) as used i n t h i s t e x t
r e f e r s t o the s t a n d a r d d e v i a t i o n o f the " f i n a l " s i g n a l ; i f s i g n a l
a v e r a g i n g o r l e a s t squares f i t t i n g i s employed t o a r r i v e a t the
f i n a l s i g n a l , then t h i s s h o u l d be i n t e r p r e t e d as the s t a n d a r d
error.]
3.1.4 non-white n o i s e . The a u t o c o r r e l a t i o n f u n c t i o n ( o r
s p e c t r a l power d e n s i t y ) must be t a k e n i n t o account i n c a l c u l a t i n g
c r i t i c a l l e v e l s or confidence i n t e r v a l s . This i s not a t r i v i a l
m a t t e r , and i s remarkably o f t e n i g n o r e d . I t s importance i s seen
most o f t e n f o r time dependent phenomena [e.g., i n chromatography],
and where " f l i c k e r n o i s e " i s found. Note t h a t n o i s e o f t h i s s o r t
s e t s l i m i t s t o the g a i n s w h i c h may be a c h i e v e d through signal
i n t e g r a t i o n o r a v e r a g i n g . Note a l s o t h a t d e t e c t i o n l i m i t s based on
the S i g n a l t o Background R a t i o d e r i v e from the a s s u m p t i o n o f
b a c k g r o u n d - c a r r i e d f l i c k e r n o i s e dominance. See e s p e c i a l l y Smit
(40) and E p s t e i n (41) f o r i m p o r t a n t d i s c u s s i o n s o f t h i s t o p i c . In
a b r o a d e r sense the u n d e r l y i n g i s s u e r e l a t e s t o the l i m i t i n the
i n f o r m a t i o n c o n t e n t o f s e t s o f o b s e r v a t i o n s w h i c h a r e not f u l l y
independent. One encounters i t a l s o when i n t e r p r e t i n g u n c e r t a i n -
t i e s f o r count r a t e meters (RC s i g n a l averages - (2J3, p. 9 6 ) ) , and
u n c e r t a i n t i e s i n f u n c t i o n s o f p a r t i a l l y c o r r e l a t e d random v a r i a b l e s
[error propagation, including covariance: (42)]. The t o p i c i s o f
s p e c i a l r e l e v a n c e when c o n s i d e r i n g i n s t r u m e n t a l b a s e l i n e s [see
below].

3.1.5 non-normal d i s t r i b u t i o n s . T h i s problem can be d e a l t

w i t h r i g o r o u s l y i f one knows the form o f the random e r r o r d i s t r i b u -
tion. A n o t a b l e example o f t h i s o c c u r s i n " c o u n t i n g " experiments
( e . g . , r a d i o a c t i v i t y ) , where the p h y s i c s o f the p r o c e s s i m p l i e s
Poisson s t a t i s t i c s . As the P o i s s o n d i s t r i b u t i o n i s d i s c r e t e , y c
(the c r i t i c a l l e v e l f o r g r o s s c o u n t s ) t a k e s on i n t e g e r v a l u e s o n l y ,
and a i s g e n e r a l l y i n the form o f an i n e q u a l i t y -- i e , a < 0.05
(.28) . D i s t r i b u t i o n - f r e e t e c h n i q u e s , e s p e c i a l l y those based on
o r d e r s t a t i s t i c s (such as the median and i t s c o n f i d e n c e i n t e r v a l ) ,
and t r a n s f o r m a t i o n t e c h n i q u e s (eg, f o r l o g - n o r m a l l y d i s t r i b u t e d
e r r o r s ) , a r e o f t e n a p p r o p r i a t e (26.57). S o - c a l l e d non-parametric
t e c h n i q u e s -- the Gauss or Chebyshev i n e q u a l i t i e s , g i v e ( 2 - s i d e d )
2 2
a's as no g r e a t e r than ( 2 / [ 3 k ] ) and 1/k r e s p e c t i v e l y , where the
s t a n d a r d d e v i a t i o n m u l t i p l i e r k r e p l a c e s z o f the normal d i s t r i b u -
tion. Note t h a t the Gauss I n e q u a l i t y i s a p p l i c a b l e f o r random
v a r i a b l e s h a v i n g unimodal continuous and symmetric d e n s i t
f u n c t i o n s , whereas the
d i s t r i b u t i o n having f i n i t
a p p l y i n g the i n e q u a l i t i e s i s t h a t k must m u l t i p l y a, and a i s n o t
2 2
g e n e r a l l y known. A l t h o u g h s i s an u n b i a s e d e s t i m a t e f o r a even
for non-normal d i s t r i b u t i o n s , bounds f o r s/a a r e d i s t r i b u t i o n
dependent and t h e r e f o r e a l s o not g e n e r a l l y known. fNote 4. ]
D i f f i c u l t i e s are compounded when the measurement p r o c e s s c o n s i s t s
of two o r more s t e p s c o m p r i s i n g d i f f e r e n t k i n d s o f p d f ' s . [See f o r
example, Johnson, Ref. (44).]
Recommended s o l u t i o n s f o r the n o n - n o r m a l i t y problem a r e : 1)
use the percentage p o i n t s o f the a c t u a l p d f , i f known; b) t r a n s f o r m
to n o r m a l i t y ; c) use o r d e r s t a t i s t i c s ; d) d e s i g n the experiment t o
t a k e advantage o f " p a i r i n g " and the C e n t r a l L i m i t Theorem. The
l a s t approach, which l o o k s v e r y a t t r a c t i v e f o r c h e m i c a l r e s e a r c h ,
w i l l be d i s c u s s e d below. I n f o r m a t i o n on the o t h e r approaches may
be o b t a i n e d from s p e c i a l i z e d s t a t i s t i c a l t e x t s ( 4 5 ) .
3.1.6 p a i r e d comparisons: C e n t r a l L i m i t Theorem. The C e n t r a l
L i m i t Theorem makes q u a l i t y c o n t r o l c h a r t s work. Here, one c h a r t s
s e t s o f averages o f o b s e r v a t i o n s and checks f o r e x c u r s i o n s beyond
Normal c o n t r o l l i m i t s . The a v e r a g i n g i s done n o t p r i m a r i l y f o r
s t a n d a r d e r r o r r e d u c t i o n , b u t t o a s s u r e (approximate) n o r m a l i t y .
I t can be shown t h a t averages ( o r sums) d e r i v e d from o f a sequence
of mutually independent random variables having a common
d i s t r i b u t i o n t e n d toward n o r m a l i t y , o f t e n r a t h e r q u i c k l y (by the
time n = 3 o r 4 ) . T h i s " C e n t r a l L i m i t Theorem" i s v a l i d r e g a r d l e s s
of the shape o f the i n i t i a l d i s t r i b u t i o n , so l o n g as i t has f i n i t e
variance. The r a t e o f approach t o n o r m a l i t y , however, depends on
the i n i t i a l shape, b e i n g f a s t e r f o r symmetric d i s t r i b u t i o n s ( 4 5 ) .
For l o w - l e v e l c h e m i c a l measurements, a l l too o f t e n the b l a n k , which
forms the basis for the detection decision, i s neither
s y m m e t r i c a l l y nor n o r m a l l y d i s t r i b u t e d -- e s p e c i a l l y when the b l a n k
i s due t o e n v i r o n m e n t a l or p a r t i c u l a t e c o n t a m i n a t i o n ( 4 6 ) . Very
wrong t r a c e a n a l y t i c a l c o n f i d e n c e i n t e r v a l s and d e t e c t i o n d e c i s i o n s
may r e s u l t . By d e s i g n i n g the measurement p r o c e s s so t h a t p r o p e r
p a i r e d comparisons can be made (45, Chapt. 4 ) , one can a t the same
time a c h i e v e the b e s t s t a t i s t i c a l s e n s i t i v i t y and f o r c e symmetry
(for the e s t i m a t e d n e t s i g n a l ) , and thus s e t the stage f o r
approximate n o r m a l i t y f o r averages o f such e s t i m a t e s . Two o t h e r
major reasons f o r f o r c i n g symmetry i n t h i s way a r e : a) t o take

advantage o f the median and i t s c o n f i d e n c e i n t e r v a l f o r r o b u s t

e s t i m a t i o n , and b) t o make p o s s i b l e the use o f the Gauss I n e q u a l i t y
for d i s t r i b u t i o n free interval estimation. T h i s approach has been
s u g g e s t e d , f o r example, as a p o s s i b l e s o l u t i o n t o the s e v e r e
d e t e c t i o n d i s c r e p a n c i e s o b t a i n e d i n IAEA i n t e r c o m p a r i s o n s (35).
S p e c i f i c a l l y , the recommendation i s t o make d e t e c t i o n d e c i s i o n s by
comparing the averages o f a t l e a s t n-4 p a i r e d comparisons ( g r o s s
s i g n a l - e q u i v a l e n t b l a n k s i g n a l ) w i t h t s / 7 n , where s 2
i s the
e s t i m a t e d v a r i a n c e o f the n n e t s i g n a l s . (Note the d i r e c t a n a l o g y
between the c h e m i c a l b l a n k measurement, and the " c o n t r o l " o b s e r v a -
t i o n w h i c h p l a y s a c e n t r a l r o l e i n c l i n i c a l and p s y c h o l o g i c a l n u l l
hypothesis t e s t i n g . ) I n a n t i c i p a t i o n o f the n e x t two t o p i c s , two
r e l a t e d advantages o f p r o p e r p a i r i n g ( o r e q u i v a l e n t f i t t i n g ) may be
s t a t e d : a) b i a s a s s o c i a t e d w i t h s y s t e m a t i c B-changes i s m i n i m i z e d
by t a k i n g " l o c a l " d i f f e r e n c e s (y - B) and u s i n g l o c a l v a l u e s f o r
recovery and instrumental d e t e c t i o n e f f i c i e n c y ; b) e f f e c t s o f
imprecision associated
( e . g . , "between-day") ma
3.1.7 background, b a s e l i n e , b l a n k . The v a r i a b i l i t y o f the
n u l l s i g n a l [B] i s the d e t e r m i n i n g f a c t o r i n making v a l i d d e t e c t i o n
d e c i s i o n s (H t e s t s ) o r i n d e r i v i n g v a l i d c o n f i d e n c e i n t e r v a l s f o r
Q
l o w - l e v e l s i g n a l s . I n the i d e a l i n t e r f e r e n c e f r e e , "pure s o l u t i o n "

measurement environment, the instrumental background is the
u l t i m a t e l i m i t i n g f a c t o r . I n t h i s sense an " i n s t r u m e n t a l c r i t i c a l
l e v e l " ( d e c i s i o n l e v e l , t h r e s h o l d ) and the c o r r e s p o n d i n g d e t e c t i o n
l i m i t mark the b e s t p o s s i b l e performance o f a system. Two cautions
a r e i n o r d e r , however. F i r s t , the i n s t r u m e n t a l n o i s e (background
v a r i a b i l i t y ) may not be w h i t e -- i . e . , t h e r e may be l o n g - or
s h o r t - t e r m v a r i a t i o n s w h i c h must be compensated f o r by a p p r o p r i a t e
m o d e l i n g and/or a s t u t e p a i r e d comparisons. Second, when the
c h e m i c a l , p h y s i c a l , o r g e o m e t r i c c o n f i g u r a t i o n o f the f i n a l sample
changes, t h e r e may be c o r r e s p o n d i n g changes i n the e f f e c t i v e
b a c k g r o u n d ( f o r example due t o changes i n e x t e r n a l s c a t t e r e d
radiation). I n s t r u m e n t a l d e t e c t i o n e f f i c i e n c i e s o r r e s p o n s e s may
a l s o be p e r t u r b e d by such s a m p l e - r e l a t e d f a c t o r s ( 4 7 ) , but t h a t i s
a " c a l i b r a t i o n " m a t t e r , t o be t a k e n up s e p a r a t e l y .
Multicomponent instrumental responses, unless totally
s e l e c t i v e , generate s p e c t r a l or chromatographic b a s e l i n e s which
a r i s e from complex p h y s i c o c h e m i c a l phenomena r a n g i n g from m u l t i p l e
(sample) p a r t i c l e o r r a d i a t i o n s c a t t e r i n g t o component t a i l i n g ,
depending on the s p e c i f i c a n a l y t i c a l t e c h n i q u e i n v o l v e d . Such
b a s e l i n e s g e n e r a l l y subsume any i n s t r u m e n t a l background, and thus
become the l i m i t i n g f a c t o r . V a l i d n e t s i g n a l e s t i m a t e s and detec-
tion decisions t h e n become c r i t i c a l l y dependent on accurate
m o d e l i n g o f the b a s e l i n e f u n c t i o n a l shape and n o i s e s t r u c t u r e .
E m p i r i c a l b a s e l i n e shape models are common, the l i n e a r model b e i n g
most used. D e v i a t i o n s from l i n e a r i t y may be modeled u s i n g low
o r d e r p o l y n o m i a l s o r s p l i n e s , b u t s i n c e the m o d e l i n g i s e m p i r i c a l
one must be a l e r t t o p o s s i b l e model e r r o r s , such as u n a n t i c i p a t e d
f i n e s t r u c t u r e (48). The n o i s e s t r u c t u r e o f the e m p i r i c a l l y
modeled b a s e l i n e d e s e r v e s s p e c i a l a t t e n t i o n i n the case o f d r i f t ,
p r e f e r r e d p e r i o d i c i t y , o r more g e n e r a l a u t o c o r r e l a t i o n such as "1/f
n o i s e " (40.41.49).
For a l l r e a l c h e m i c a l measurements, the c h e m i c a l b l a n k i s the
actual limiting factor. To a s s e s s i t s magnitude and v a r i a b i l i t y ,

t h e r e i s no b e t t e r approach t h a n t o a p p l y the e n t i r e measurement

p r o c e s s t o an adequate number o f r e a l b l a n k s . Unfortunately, this
i d e a l may n o t be r e a l i z a b l e , as i t r e q u i r e s samples w h i c h are
i d e n t i c a l t o those o f i n t e r e s t i n a l l r e s p e c t s e x c e p t f o r the
absence o f the t a r g e t a n a l y t e . The a l t e r n a t i v e approach i s t o
a t t e m p t t o "propagate" the components o f the b l a n k f o r each s t e p o f
the CMP, t a k i n g i n t o account the p o i n t s o f i n t r o d u c t i o n , and
subsequent r e c o v e r i e s and CMP-induced v a r i a t i o n s . This t o p i c i s
enormously i m p o r t a n t and enormously complex. One must c o n s i d e r
simultaneously: the e f f e c t s o f m u l t i p l e b l a n k s o u r c e s ; a n a l y t e ,
blank, and interferant recoveries f o r each CMP step; and
i n s t r u m e n t a l d e t e c t i o n e f f i c i e n c i e s p l u s m a t r i x e f f e c t s f o r each
(35. 50-52). A s m a l l c o m p l i c a t i o n a r i s e s from the f a c t t h a t
d i f f e r e n t types o f b l a n k s may e x h i b i t d i f f e r e n t p d f ' s . Reagent and
sample p r e p a r a t i o n b l a n k s t e n d t o be n o r m a l l y d i s t r i b u t e d , w h i l e
e n v i r o n m e n t a l b l a n k s ar f r e q u e n t l l o g - n o r m a l (44.46.53) I th
f i n a l a n a l y s i s , of course
convolved w i t h instrumenta
n o r m a l i t y f o r e i t h e r w i l l be r e f l e c t e d i n the f i n a l , e f f e c t i v e
blank d i s t r i b u t i o n . A f i n a l observation: s y s t e m a t i c o r random
e r r o r i n the e s t i m a t e d b l a n k a f f e c t s n o t o n l y d e t e c t i o n l i m i t s and
c o n f i d e n c e i n t e r v a l s f o r " l o w - l e v e l " samples; i t may a l s o l i m i t
the a c c u r a c y o f h i g h p r e c i s i o n , " h i g h - l e v e l " samples.
3.1.8 internal vs external error: propagation vs
replication. The u n c e r t a i n t i e s o f low l e v e l concentration
e s t i m a t e s may be d e r i v e d from e r r o r p r o p a g a t i o n f o r each s t a g e o r
s t e p o f a compound CMP, o r they may be deduced from r e p l i c a t i o n
and comparison w i t h l o w - l e v e l SRMs f o r the o v e r a l l measurement
p r o c e s s , as i n l a b o r a t o r y i n t e r c o m p a r i s o n s . C o n s i s t e n c y between
the two approaches i s e s s e n t i a l f o r the u n c e r t a i n t y e s t i m a t e s t o be
considered v a l i d . Among the e r r o r c h a r a c t e r i s t i c s t h a t may be
exposed through such ANOVA type t e s t i n g a r e : " e x c e s s " random e r r o r
[ a ] , s y s t e m a t i c e r r o r [ A ] , and c o v a r i a n c e among i n t e r n a l e r r o r s .
x
The f i r s t (^ ) x might r e p r e s e n t , f o r example, a between day o r

between l a b v a r i a n c e component, o r i n the case o f P o i s s o n c o u n t i n g
s t a t i s t i c s , an excess o r n o n - c o u n t i n g component o f random e r r o r .
The second (A) c o u l d be m a n i f e s t as the d i f f e r e n c e between the
l i m i t i n g mean f o r an i n t r a l a b o r a t o r y measurement and the t r u e
( e . g . , SRM) v a l u e . The t h i r d might be seen i f i n t e r n a l , non-white
n o i s e o r o t h e r e r r o r c o r r e l a t i o n e f f e c t s were i m p r o p e r l y a c c o u n t e d
f o r i n e r r o r propagation; the e x t e r n a l e s t i m a t e , d e r i v e d from
independent r e p l i c a t e s would a u t o m a t i c a l l y compensate f o r such
( i n t e r n a l ) behavior. Comparison o f i n t e r n a l and e x t e r n a l e s t i m a t e s
has i t s l i m i t a t i o n s , however. T h i s i s shown i n F i g . 6 [ ( a , A ) vs x D
n] w h i c h d i s p l a y s the, d e t e c t i o n l i m i t s f o r a and A as a f u n c t i o n
x
o f the number o f r e p l i c a t e s (54)- Thus, i n o r d e r t o d e t e c t b i a s

e q u a l i n magnitude t o the s t a n d a r d d e v i a t i o n , one needs a t l e a s t 12
degrees o f freedom (13 r e p l i c a t e s ) . To d e t e c t an e x t r a v a r i a n c e
component e q u a l i n magnitude t o the known i n t e r n a l p r e c i s i o n
[ ( a ) - a ] , one needs 46 degrees o f freedom ( 5 5 ) . These e r r o r com-
x D i
ponents [ a , A] must, however, be t a k e n i n t o a c c o u n t i f v a l i d

x
d e t e c t i o n d e c i s i o n s are t o be made.

F i g . 6. D e t e c t i o n L i m i t s f o r B i a s [A] and excess ( e x t e r n a l )

random e r r o r [a ] v s t h e Number o f O b s e r v a t i o n s .
x (Adapted from
Ref. 54.)

3.1.9 I n t e r n a l e r r o r p r e c i s e l y known: Improve^ fletecfripn

feclgJQTlg ((MIA c o n f l u e n c e i n t e r v a l s ) y s l n g I n e q u a l i t y c o n s t r a i n t s ,
I f t h e excess random e r r o r component i s w e l l known, t h e n o b v i o u s l y
e r r o r p r o p a g a t i o n c a n be a p p l i e d -- e.g., V - V + Vfc -- t o
x
c a l c u l a t e t h e t o t a l v a r i a n c e V , w h i c h i s t h e q u a n t i t y t h a t must be
t
used f o r c a l c u l a t i n g the c r i t i c a l l e v e l o r confidence i n t e r v a l .

(For n o t a t i o n a l s i m p l i c i t y , V i s used h e r e t o denote v a r i a n c e , i n
2
p l a c e o f a . I f m u l t i p l i c a t i v e r a t h e r than a d d i t i v e r e l a t i o n s are
i n v o l v e d V w o u l d r e p r e s e n t r e l a t i v e v a r i a n c e . ) Q u i t e f r e q u e n t l y we
f i n d i t r e l a t i v e l y inexpensive t o o b t a i n p r e c i s e estimates o f V i
( i n t e r n a l v a r i a n c e ) whereas e x t e r n a l r e p l i c a t e s -- e.g., between

l a b o r a t o r i e s o r between days, e t c . , t e n 4 t o be more c o s t l y ,
hence fewer. Perhaps t h e extreme case o f t h i s s o r t o c c u r s w i t h
P o i s s o n c o u n t i n g s t a t i s t i c s , where V N where N i s t h e observed
i
number o f c o u n t s . I f expected value o f N i s s u f f i c i e n t l y l a r g e

( e . g . , >60) t h e e x p r e s s i o n y i e l d s a r e a s o n a b l y good e s t i m a t e f o r
V , as good as w o u l d r e s u l
i
t a k e n t o be z e r o , t h
d i r e c t l y from V w h i c h h e r e i s e s t i m a t e d as t h e number o f counts
i
fNote 51.
If V x i s n o t known, we have t h r e e a l t e r n a t i v e s . One o f
extreme c o n s e r v a t i s m would be t o use t h e lower and upper l i m i t s f o r
2
V , b a s e d on r e p l i c a t i o n
x [ s - e s t ( V ) ] and knowledge o f V .
t L
I n t e r m e d i a t e , and most common, i s s i m p l y t o c a l c u l a t e as t * s ,

where t i s b a s e d on t h e e x t e r n a l number o f d f (number o f r e p l i c a t e s
minus o n e ) . More i n t e r e s t i n g i s t h e use d f o u r knowledge t h a t
V >0, and a v a r i a n c e w e i g h t e d t . (A f o u r t h a l t e r n a t i v e , i g n o r i n g
x
the p o s s i b l e e x i s t e n c e o f V xi s a l l t o o common; t h e unsupported

assumption t h a t counting s t a t i s t i c s o r other i n t e r n a l i n s t r u m e n t a l
v a r i a n c e f i x e s t h e o v e r a l l i m p r e c i s i o n c a n generate T-^'s and C I ' s
t h a t a r e t o o s m a l l and c o r r e s p o n d i n g l y l a r g e f a l s e positive
probabilities.)
The m e r i t o f t h e v a r i a n c e w e i g h t e d t t e c h n i q u e i s t h a t i t
p e r m i t s us t o use o u r e x c e l l e n t knowledge o f together w i t h the
f a c t t h a t V cannot be n e g a t i v e t o o b t a i n a s i g n i f i c a n t l y s m a l l e r
x
2
o r CI than we would u s i n g s d i r e c t l y . I t provides p r o t e c t i o n
a g a i n s t u n a n t i c i p a t e d e x t e r n a l random e r r o r w i t h l i t t l e p e n a l t y i f
t h a t e r r o r component i s i n f a c t n e g l i g i b l e . The t e c h n i q u e sug-
g e s t e d h e r e i s t e n t a t i v e and approximate, b u t i t appears t o be
c o n s e r v a t i v e and a s y m p t o t i c a l l y c o r r e c t fNote 61.
To i l l u s t r a t e , l e t us c o n s i d e r t r i p l i c a t e measurements o f a
sample u s i n g a c o u n t i n g t e c h n i q u e , such as i o n c o u n t i n g mass
s p e c t r o m e t r y o r photon c o u n t i n g i n o p t i c a l s p e c t r o m e t r y , X - r a y
f l u o r e s c e n c e a n a l y s i s o r gamma r a y s p e c t r o m e t r y . Internal variance
d e r i v e s from P o i s s o n c o u n t i n g s t a t i s t i c s [V ] where t h e a p p r o p r i a t e

v a l u e o f t e q u a l s i t s normal l i m i t z
L o r 1.645 f o r a - 0.05.

T o t a l v a r i a n c e [ V ] f o r t h e 3 r e p l i c a t e s i s e s t i m a t e d as s , where
t
2
tt i s 2 . 9 2 f o r 2 d f . Excess v a r i a n c e [V ] i s V -V , and e s t i m a t e d
x t L
2
as s - V , w i t h t h e c o n s t r a i n t t h a t V may n o t be n e g a t i v e . 1^,' ( o r
i x
CI) i s c a l c u l a t e d as t's'/Jri where:
f t (V./Vt) + t t (V A ) x t (2)

2
F o r the example a t hand, we e s t i m a t e V x as s - V .
i (Eq. 1) thus
y i e l d s s'=a i i f s < , o r s'=s i f s > a . (Eq. 2) becomes
2 2
t'=1.645 (k) + 2.92 (1-k) where k - a . / s . For example, i f a. i s
e q u i v a l e n t t o 1.75 ng-Ca, and s, t o 3.04 ng-Ca, k would e q u a l
2
( 1 . 7 5 / 3 . 0 4 ) = 0.331. As a r e s u l t , s' - 3.04 ng-Ca, t ' - 2.50, and
1^' = 4.39 ng-Ca. I n v e s t i g a t i o n o f the p r o p e r t i e s o f ' f o r V =0
x
to V V shows i t t o be c o n s e r v a t i v e [a' < 0.05] w i t h a l i m i t i n g

x A
v a l u e when V==0 o f a p p r o x i m a t e l y 0.03.

x A l s o i n t h i s l i m i t i n g case,
1^, ' on the average i s o n l y s l i g h t l y g r e a t e r (< 10%) t h a n i t would
be i f one assumed V was i d e n t i c a l l y z e r o , whereas f o r the conven-
x
t i o n a l approach [1^ - 2.92(s/7n)) i t would be 78% l a r g e r . When the

s t a k e s a r e h i g h e r -- e.g., CI's o r L^'s f o r a=0.01 -- the c o n t r a s t
becomes even g r e a t e r . I n e f f e c t , we have used our knowledge o f V i
to e x c l u d e v e r y s m a l l v a l u e s f o r e s t i m a t e d t o t a l a, and g a i n e d
smaller CI's, ' s , and d e t e c t i o n l i m i t s i n r e t u r n .
3.1.10 e f f e c t s o f r o u n d i n g and truncation Premature
rounding of experimenta
r e s u l t i n g i n erroneou
distribution, i t s parameters [mean, v a r i a n c e ] , and r e s u l t s o f
s t a t i s t i c a l t e s t s ( e . g . , d e t e c t i o n d e c i s i o n s , q u a l i t y o f f i t ) and
confidence i n t e r v a l s . The most o b v i o u s d i s t o r t i o n i s t h a t an
i n h e r e n t l y c o n t i n u o u s d i s t r i b u t i o n i s made d i s c r e t e ; the e f f e c t i s
analogous t o " d i s c r e t i z a t i o n n o i s e " w h i c h i s o f t e n found w i t h
m u l t i c h a n n e l and m u l t i d e t e c t o r a r r a y t e c h n i q u e s i n v o l v i n g windows
i n t i m e , space, energy, w a v e l e n g t h , e t c . ( 5 6 ) . The tolerable
degree o f r o u n d i n g depends on the d i s t r i b u t i o n . For n o r m a l l y
d i s t r i b u t e d d a t a , t h e r e i s about a 10% chance o f f i n d i n g r e s u l t s
w i t h i n a/8 o f the mean. S c a l e d i v i s i o n s much s m a l l e r t h a n a/4 a r e
t h e r e f o r e r e q u i r e d i f one i s t o a v o i d f a l s e c o i n c i d e n c e s , and f i t s
t h a t a r e "too good", e t c . I n f a c t , c l u e s t o e x c e s s i v e r o u n d i n g o r
2
t r u n c a t i o n may be found i n x o r
F s t a t i s t i c s which are u n u s u a l l y
s m a l l , o r i n p d f ' s e x h i b i t i n g unexpected d e v i a t i o n s from n o r m a l i t y
(57) . A b n o r m a l i t y i s n o t e d a l s o by Cheeseman and W i l s o n f o r
c o n s t r a i n e d b a l a n c e - p o i n t measurements, such as the galvanometer
needle which i s p h y s i c a l l y c o n f i n e d to non-negative s c a l e readings
(36). The importance of these c o n s i d e r a t i o n s f o r databases
i n c o r p o r a t i n g l o w - l e v e l r e s u l t s i s discussed i n (34).
3.1.11 the e v a l u a t i o n p r o c e s s [ d a t a r e d u c t i o n : fitting].
The d a t a e v a l u a t i o n p r o c e s s [EP] i s an i n t e g r a l p a r t o f the CMP,
and as such i t h e l p s d e f i n e o and the c r i t i c a l l e v e l .
Q It is
perhaps o b v i o u s t h e n t h a t 1^, , C I ' s , and the d e t e c t i o n l i m i t w i l l
d i f f e r f o r the v e r y same e x p e r i m e n t a l d a t a , depending on the EP
a p p l i e d . A s i m p l e i l l u s t r a t i o n i s found i n the f i t t i n g o f s p e c t r a l
or c h r o m a t o g r a p h i c peaks. One may use the peak h e i g h t as the
q u a n t i t a t i v e s i g n a l measure, o r a model-independent peak a r e a may
be used, o r a more s o p h i s t i c a t e d t e c h n i q u e such as l i n e a r o r
n o n - l i n e a r l e a s t squares may be employed t o e s t i m a t e the peak s i z e
a c c o r d i n g t o a s e l e c t e d f u n c t i o n a l model such as a G a u s s i a n o r
skewed G a u s s i a n (58). The point i s that without explicit
s p e c i f i c a t i o n o f the e n t i r e CMP, i n c l u d i n g t h e EP employed, the
d e t e c t i o n c h a r a c t e r i s t i c s o f the measurement p r o c e s s a r e u n d e f i n e d .
Because o f t h i s , a s l i g h t problem o c c u r s when the EP i s g i v e n as a
" b l a c k box", o r a l g o r i t h m whose c h a r a c t e r i s t i c s a r e u n c l e a r . ( T h i s
i s s u e , i n c l u d i n g the common a v a i l a b i l i t y o f e x e c u t a b l e s o f t w a r e

w i t h o u t source code, w i l l be t r e a t e d f u r t h e r i n the d i s c u s s i o n o f

d e t e c t i o n l i m i t s i n the n e x t s e c t i o n . )
When the EP comprises l i n e a r computations ( l i n e a r i n the
o b s e r v a t i o n s ) such as s i m p l e d i f f e r e n c e s , y - B, or l i n e a r l e a s t
squares or l i n e a r m u l t i v a r i a t e c o m p u t a t i o n s , i n i t i a l n o r m a l i t y ( o f
the o b s e r v a t i o n s y) i s p r e s e r v e d f o r the e s t i m a t e d q u a n t i t i e s .
N o n - l i n e a r c o m p u t a t i o n s , such as a r i s e commonly i n i t e r a t i v e model
s e l e c t i o n and peak s e a r c h r o u t i n e s , produce e s t i m a t e d parameters
h a v i n g non-normal d i s t r i b u t i o n s (.59). Caution i s i n order, i n
t h o s e c a s e s , i n a p p l y i n g "normal" v a l u e s o f t e s t s t a t i s t i c s t o
c a l c u l a t e 1^-. and C I ' s . (Other f a c t o r s t o c o n s i d e r are the e x t e n t
o f n o n - l i n e a r i t y , the l e v e l o f c o n f i d e n c e or s i g n i f i c a n c e [ 1 - a ] ,
and the r o b u s t n e s s o f the s t a t i s t i c i n q u e s t i o n . )
F i n a l l y , i t s h o u l d be n o t e d t h a t an erroneous model w i l l g i v e
erroneous r e s u l t s . T h i s seeming t r u i s m i s i m p o r t a n t because models
2
w h i c h pass s t a t i s t i c a l t e s t [e.g. x test f f i t ] consistent
w i t h the d a t a b u t not n e c e s s a r i l
linearity, model erro
s i g n i f i c a n t b i a s i n the r e s u l t s (48).
3.1.12 c a l i b r a t i o n error. A number o f d i f f e r e n t approaches
may be t a k e n t o i n c o r p o r a t e the u n c e r t a i n t y i n the c a l i b r a t i o n
f a c t o r A i n t o the c r i t i c a l l e v e l . To i l l u s t r a t e , l e t us c o n s i d e r
the s i m p l e s t f u n c t i o n a l r e l a t i o n f o r the E v a l u a t i o n P r o c e s s :
x - (y - B)/A - S/A (3)
U n f o r t u n a t e l y , t h i s i s a l r e a d y a n o n - l i n e a r r e l a t i o n , so we cannot
e x p e c t x t o be n o r m a l l y d i s t r i b u t e d . I f the r e l a t i v e e r r o r i n A i s
s m a l l (e.g., < 10%) i t s i n f l u e n c e on 1^, i s l i k e w i s e s m a l l , and
d e v i a t i o n s from n o r m a l i t y are m i n i m a l . I f the r e l a t i v e u n c e r t a i n t y
i n A i s not n e c e s s a r i l y s m a l l , or i f i t i n c l u d e s possible
s y s t e m a t i c e r r o r , a s t r a i g h t f o r w a r d approach i s t o use the lower
bound f o r A t o c a l c u l a t e an upper bound f o r L^. (here x ) w h i c h can c
be u s e d t o make c o n s e r v a t i v e detection decisions [a < 0.05].

( I n c o r p o r a t i o n o f bounds f o r s y s t e m a t i c e r r o r i s d i s c u s s e d more
f u l l y i n the s e c t i o n on d e t e c t i o n l i m i t s . )
E r r o r p r o p a g a t i o n from the f i t t i n g o f a c a l i b r a t i o n c u r v e can
be used to t r e a t d e t e c t i o n and interval estimation (almost)
r i g o r o u s l y p r o v i d e d the model i s c o r r e c t -- a c a u t i o n b e i n g t h a t
the i n t e r c e p t - B may not represent the b l a n k - B (60.61). An
i n t e r e s t i n g a l t e r n a t i v e i s to estimate a Q and the c o r r e s p o n d i n g
detection characteristics directly for x [i.e., i n units of
c o n c e n t r a t i o n ] by f u l l r e p l i c a t i o n o f the CMP a t the l e v e l s o f
c o n c e r n , o b s e r v i n g y, B, and A f o r each r e p l i c a t e . ( T h i s i s the
" p a i r e d comparison" concept extended t o c a l i b r a t i o n , where a b l a n k
and s t a n d a r d i s r u n f o r e v e r y sample.) The s t a t i s t i c a l p r o p e r t i e s
o f the o b s e r v e d x d i s t r i b u t i o n can t h e n be used t o d i r e c t l y
calculate x c [as t s , i f A - v a r i a t i o n i s not too g r e a t ] and e s t i m a t e
Q
the d e t e c t i o n l i m i t . An added b e n e f i t o f t h i s scheme i s t h a t

d i r e c t o b s e r v a t i o n o f the b l a n k decouples i t from the c a l i b r a t i o n
curve f i t t i n g process, so t h a t an assumed s t r a i g h t l i n e model
[ c o n s t a n t s e n s i t i v i t y A] can be t e s t e d by f i t t i n g a l i n e [S=Ax+e]
t h r o u g h the o r i g i n (62.). For v a l i d c o n c l u s i o n s , o f c o u r s e , due
a t t e n t i o n must be g i v e n to i n t e r f e r e n c e and m a t r i x e f f e c t s on b o t h

parameters, B and A. An i l l u s t r a t i o n of an observed x d i s t r i b u t i o n

1 3 1
for the n u l l hypothesis [x - 0] i s shown f o r I i n (44).
3.1.13 reporting of l o w l e v e l data. Problems associated with
data rounding and truncation extend to the reporting of f i n a l
r e s u l t s . Also, j u s t as i n the case of the data evaluation step of
the CMP, reporting must be treated as an i n t e g r a l part of the
o v e r a l l CMP. Bias and information-loss are the prime considera-
tions. At the lower extreme, where x - 0 [null hypothesis]
suppression of negative estimates forces a p o s i t i v e bias, on the
average. Other biases a r i s e when a l l non-detected r e s u l t s are
reported as zero or as equal to (or less than) the detection l i m i t .
The d i f f i c u l t i e s are evident as soon as one attempts to: develop a
database comprising large amounts of low-level data (14); to
compute temporal or s p a t i a l averages for higher order detection
decisions (2); or to compute average concentrations across
d i f f e r e n t materials as i n the USDIET-1 exercise (62) This l a s t
example i l l u s t r a t e s th
were prepared for measuremen
toxic chemical constituents, including the trace element Se.
Comparison of the r e s u l t for Se i n the composite sample [128 /ig/g]
with the weighted average from the large number of i n d i v i d u a l
contributing foods [100 Mg/g] showed a s i g n i f i c a n t negative bias
for the l a t t e r . This was a r e s u l t of s e t t i n g a l l "trace" observa-
tions (defined as those below a q u a n t i f i c a t i o n l i m i t , L^) to zero.
Adjusting these upward to L^/2 led to an improvement [110 Mg/g]
but negative bias was s t i l l apparent. These kinds of problems can
be completely circumvented i f concentration estimates, even i f
negative, are always reported together with t h e i r uncertainties
(64. 65) . Detection decisions can be made by comparison with ,
and upper l i m i t s may be given as x + Zs/Jri.
3.1.14 thresholds. The threshold for discriminating "real
signals" from blanks may be set i n various ways. The only way that
i s consistent with the r e l a t i o n between confidence intervals and
significance tests i s the one described, Lg - t s . Other techniques
include the use of a constant m u l t i p l i e r ks or ka, with k - 3 a
popular choice; and use of a f i x e d threshold signal or concentra-
tion, such as 1 mV or 2 ng. A drawback of these alternative
techniques i s that they seldom recognize the existence or magnitude
of the a-error, which, however, does e x i s t , and which w i l l take on
varying values depending on the number of degrees of freedom or the
magnitude of the f i x e d threshold i n comparison to a . For con- Q
fidence intervals a i s conventionally taken as 0.05, so there seems

l i t t l e j u s t i f i c a t i o n for depressing i t by a factor of f o r t y to
0.0013 (corresponding to 3a) for detection decisions [a-known], or
by more than a factor of ten (corresponding to 3s) for 20 degrees
of freedom. Instruments having hardware or software discriminators
may have r e s u l t i n g dead zones which correspond to vanishingly small
a's. In one case recently, the threshold was set so high that a
was beyond the range of any of the s t a t i s t i c a l tables, 1^/^ 34;
i n fact the threshold was so high that the c r i t i c a l l e v e l exceeded
even the conventional l i m i t of q u a n t i f i c a t i o n -- i . e . , the RSD at
Lg was but 5% (61, p. 76, case-e) ! Such high and varying thresh-
olds for detection decisions lead both to needless confusion and to
measurement processes operating f a r short of t h e i r inherent
capabilities.

3.2 M a t t e r s C o n c e r n i n g the E r r o r o f the Second K i n d , and the

A n a l y t e D e t e c t i o n L i m i t . The c o n c e r n i n the p r e c e d i n g s e c t i o n was
the v a l i d i t y o f d e t e c t i o n d e c i s i o n s , based on comparisons o f
e x p e r i m e n t a l outcomes [ e s t i m a t e d s i g n a l s o r c o n c e n t r a t i o n s ] w i t h
a p p r o p r i a t e c r i t i c a l l e v e l s o r d e c i s i o n t h r e s h o l d s . Here, we t u r n
to i s s u e s c o n c e r n i n g the i n h e r e n t d e t e c t i o n c a p a b i l i t y o f the CMP
i n question t h a t i s , the t r u e s i g n a l s o r c o n c e n t r a t i o n s w h i c h
can be d e t e c t e d w i t h , f o r example, a 95% p r o b a b i l i t y [=0.05],
g i v e n the c r i t i c a l l e v e l ( o r e q u i v a l e n t l y a) t o b e used f o r t e s t i n g
o b s e r v e d r e s u l t s . L^ i s thus t i e d i n t i m a t e l y t o and t o a o r .
A l t h o u g h a s i g n i f i c a n c e t e s t may be p e r f o r m e d w i t h no c o n s i d e r a t i o n
o f H o r the d e t e c t i o n l i m i t , Lp i s ambiguous w i t h o u t the s p e c i f i -
A
c a t i o n o f p and a ( o r ). That i s , f o r a g i v e n Lp t h e r e i s an
i n f i n i t e s e t o f p o s s i b l e a, p p a i r s . Passing a s i g n i f i c a n c e test
-- e.g., x < x c -- i s commonly s a i d t o mean "acceptance" o f the
n u l l h y p o t h e s i s -- i . e . x 0 This i s unfortunate terminology
for only consistenc
demonstrated. " P r o o f "
bounds) demands a t t e n t i o n t o a l l p o s s i b l e a l t e r n a t i v e hypotheses
H;
A t h a t i s the t e s t i n use must be s u f f i c i e n t l y p o w e r f u l t o
" d e t e c t " [0 < 0.05, g i v e n a - 0.05] A H. A major r e a s o n f o r
i n t e r e s t i n d e t e c t i o n l i m i t s i s thus t o a l l o w us t o s e l e c t o r
d e s i g n a measurement p r o c e s s h a v i n g the c a p a c i t y t o d e t e c t s i g n a l s
o r a n a l y t e s a t p r e s c r i b e d l e v e l s o f importance. An o v e r v i e w o f
selected technical issues follows.
3.2.1 i g n o r a n c e o f the e r r o r o f the second k i n d (B). False
n e g a t i v e s o c c u r whether t h e i r e x i s t e n c e i s r e c o g n i z e d o r n o t . The
common p r a c t i c e o f making d e t e c t i o n d e c i s i o n s a t the s o - c a l l e d
d e t e c t i o n l i m i t , o r LOD, e t c . , has the e f f e c t o f s e t t i n g - 1^ ,
w i t h the r e s u l t t h a t p - 50% -- e q u i v a l e n t t o the p r o v e r b i a l f l i p
o f the c o i n . W i t h a a - c o e f f i c i e n t o f 3, a may be as s m a l l as
0.0013, r e s u l t i n g i n an imbalance [P/a] o f a f a c t o r o f n e a r l y 400!
Ignorance o f t h i s m a t t e r makes p o s s i b l e i n a d v e r t e n t o r even i n t e n -
t i o n a l m i s r e p r e s e n t a t i o n o f d e t e c t i o n c a p a b i l i t y . For example, the
s u b t l e t r a d e - o f f between a and p c o u l d be employed t o a v o i d
p e n a l t i e s f o r f a l s e p o s i t i v e s a s s o c i a t e d w i t h an inadequately
c o n t r o l l e d blank.
3.2.2 lower and upper d e t e c t i o n l i m i t s . For c e r t a i n types o f
chemical measurements there are dual null hypotheses and
c o n s e q u e n t l y d u a l 1^'s and Lp's f o r c o n c e n t r a t i o n s d i f f e r i n g from
these n u l l l e v e l s . Examples are found where a lower l i m i t i s s e t
by background n o i s e , and an upper l i m i t , by some type o f maximum
s i g n a l l i m i t a t i o n such as i n s t r u m e n t a l d e t e c t o r s a t u r a t i o n . A d u a l
i l l u s t r a t i o n i s shown i n F i g . 7 f o r two e x p o n e n t i a l phenomena,
r a d i o a c t i v e decay and r a d i a t i o n a b s o r p t i o n . I n each case the lower
Lj3 i s g i v e n by the s m a l l e s t d e t e c t a b l e d i f f e r e n c e from a comparator
( z e r o age s t a n d a r d o r b l a n k s o l u t i o n ) , and the upper Lp i s g i v e n by
the s m a l l e s t d e t e c t a b l e d i f f e r e n c e from an i n f i n i t e l y o l d sample
(no n e t e m i t t e d r a d i o a c t i v i t y ) o r an i n f i n i t e l y a b s o r b i n g sample
(no n e t t r a n s m i t t e d r a d i a t i o n .
3.2.3 the a - 8 c o n n e c t i o n : PC and ROC c u r v e s , and d e t e c t i o n
power. A c o n v e n i e n t way t o v i s u a l i z e the r e l a t i o n s h i p between
f a l s e p o s i t i v e [a] and f a l s e n e g a t i v e [p] e r r o r s and the n o r m a l i z e d
d i f f e r e n c e [d] between the means o f two p o p u l a t i o n s f o r a g i v e n
s t a t i s t i c a l t e s t has come t o us from s i g n a l d e t e c t i o n t h e o r y [ 9 ] .

F i g . 7. Lower and Upper D e t e c t i o n L i m i t s . When a measurement

p r o c e s s has b o t h minimum and maximum s i g n a l bounds, as i n r a d i o a c -
t i v e decay and o p t i c a l a b s o r p t i o n s p e c t r o m e t r y , LLD and ULD must
b o t h be c o n s i d e r e d . Dashed l i n e s i g n a l lower and upper d e t e c t i o n
l i m i t s map onto the age and c o n c e n t r a t i o n lower and upper l i m i t s
(arrows) v i a the e x p o n e n t i a l f u n c t i o n .

I n t h i s t h e o r y the R e c e i v e r O p e r a t i n g C h a r a c t e r i s t i c [ROC] curve

f o r a g i v e n t e s t t r a c e s the r e l a t i o n s h i p between the t r u e p o s i t i v e
p r o b a b i l i t y [1-0] and the f a l s e p o s i t i v e p r o b a b i l i t y [a] f o r a
g i v e n mean n o r m a l i z e d s i g n a l d i f f e r e n c e d. Another r e p r e s e n t a t i o n
o f the r e l a t i o n s h i p , denoted the Power Curve i n s t a t i s t i c s , i s the
c u r v e w h i c h t r a c e s the r e l a t i o n between D e t e c t i o n Power -- w h i c h i s
synonymous w h i c h the t r u e p o s i t i v e p r o b a b i l i t y [1-/3] -- and the
d i f f e r e n c e d, f o r a g i v e n v a l u e f o r a. (The complementary
r e l a t i o n : f} vs d, g i v e n a, i s d e s c r i b e d i n s t a t i s t i c s as the
O p e r a t i n g C h a r a c t e r i s t i c [OC] c u r v e . ) F i g . 8 shows the normal ROC
c u r v e f o r d - 3.29 i n u n i t s o f o Q [ i . e . , the d e t e c t i o n l i m i t ] , and
the power c u r v e f o r a - 0.05. The former [ROC] r e p r e s e n t a t i o n i s
the more c o n v e n i e n t f o r the comparison o f t e s t s and the s e l e c t i o n
o f a l t e r n a t i v e a, p a i r s , f o r a g i v e n d i f f e r e n c e i n p o p u l a t i o n
means. For t h i s reason, i t i s used t o compare the d i a g n o s t i c power
of a l t e r n a t i v e t e s t s i n c l i n i c a l chemistry where t h e r e are two
discrete populations [
e x a m i n i n g the v a l u e o f
p a i r i n a r e g u l a t o r y s e t t i n g where, f o r example, a s p e c i f i e d
d i f f e r e n c e [6^] i s o f c o n c e r n . A l s o , i f the s o u r c e s o r i d e n t i t i e s
o f c h e m i c a l s p e c i e s are c h a r a c t e r i z e d by unique element or i s o t o p e
r a t i o s , an ROC c u r v e c o u l d be used t o r e p r e s e n t the d i s c r i m i n a t i n g
power o f s e l e c t e d measurement t e c h n i q u e s . The second [Power c u r v e ]
r e p r e s e n t a t i o n i s more a p p r o p r i a t e when one i s i n t e r e s t e d i n the
d e t e c t i o n power as a f u n c t i o n o f ( n e t ) s i g n a l l e v e l o r c o n c e n t r a -
tion. Thus, i t i s c l e a r from the c u r v e t h a t the power i s b u t 50%
when d 1.645. A second s c a l e on the a b s c i s s a makes i t c o n v e n i e n t
in this representation to see the relation i n units of
concentration.
OC and power c u r v e s are r e g u l a r l y used i n the e v a l u a t i o n o f
statistical t e s t s (67.68). S i m i l a r l y , one f i n d s ROC curves
employed i n m e d i c i n e and p s y c h o l o g y [12, d i s c u s s i o n & r e f e r e n c e s i n
Chapt. 2 5 ] . They appear t o be l i t t l e used i n A n a l y t i c a l C h e m i s t r y ,
though L i t e a n u and R i c a have proposed the use o f d i f f e r e n t two
d i m e n s i o n a l p r o j e c t i o n s o f the t h r e e d i m e n s i o n a l r e l a t i o n s h i p [a,
ft, d] as r e p r e s e n t a t i o n s o f the " d e t e c t i o n c h a r a c t e r i s t i c " ( 8 ) .
3.2.4 power of the t-test. The three dimensional
r e l a t i o n s h i p d e s c r i b e d above i s expanded t o f o u r f o r S t u d e n t ' s t ,
w i t h the a d d i t i o n o f the number o f degrees o f freedom. I f we
r e s t r i c t our a t t e n t i o n t o the d e t e c t i o n l i m i t , by f i x i n g a and /?
b o t h t o 0.05, the r e m a i n i n g two dimensions can be v i e w e d as a
c u r v e , d vs d f -- i . e . , the d e t e c t i o n l i m i t ( i n u n i t s o f a ) as a Q
f u n c t i o n o f the number o f degrees o f freedom, where t _ i s used x a
f o r making d e t e c t i o n d e c i s i o n s . I n t h i s c a s e , the v a l u e o f d i s
d e t e r m i n e d by r e q u i r i n g a 95% p r o b a b i l i t y (1-/9) t h a t the e s t i m a t e d
n e t s i g n a l d i v i d e d by i t s e s t i m a t e d s t a n d a r d d e v i a t i o n [ ( y - B ) / s ] Q
w i l l exceed the c r i t i c a l l e v e l f o r S t u d e n t ' s t . T h i s r a t i o i s

c a l l e d the n o n - c e n t r a l t , w i t h n o n - c e n t r a l i t y parameter d, because
i t i s d i s p l a c e d from z e r o by t h i s amount. The n e t s i g n a l d e t e c t i o n
l i m i t i s g i v e n by da . Q Another i m p o r t a n t a p p l i c a t i o n o f the
non-central t d i s t r i b u t i o n has been t o t e s t the v a l i d i t y o f
presumed d e t e c t i o n l i m i t s , f o r example, i n c o n n e c t i o n w i t h m e d i c a l
d i a g n o s t i c d e v i c e s (69.) .

1.00
0.50
0.05 0.10 0.20 0.30
False Positive (a)
1.00
0.95
0.80 - [<* = 0.05]
0.60
I
0.40
0.50
A !
i 0.20
i i i
3 1.0 L C
2.0 3.0 L D
i i i
) 20 40 60 t
66
ng Ca
F i g . 8. D e t e c t i o n Power. ROC and Power ( o r OC) c u r v e s y i e l d
a g r a p h i c a l d i s p l a y o f t h e r e l a t i o n s among d e t e c t i o n limits
( d e t e c t a b l e d i f f e r e n c e s , d ) , and e r r o r s o f t h e f i r s t (a) and second
(P) k i n d s . F i g . 8A i s t h e ROC c u r v e c o r r e s p o n d i n g t o two normal
p o p u l a t i o n s d i f f e r i n g by -- i . e . , t h e s e p a r a t i o n e q u a l s 3.29 a ,
Q
and t h e c u r v e p a s s e s t h r o u g h t h e p o i n t a-/M).05. F i g . 8B i s t h e
c o r r e s p o n d i n g power c u r v e , where now a i s f i x e d , and t h e power o f
the t e s t i s g i v e n as a f u n c t i o n o f t h e n o r m a l i z e d d i s t a n c e d. The
l o w e r a b s c i s s a shows t h e e q u i v a l e n t c o n c e n t r a t i o n s c a l e f o r a
h y p o t h e t i c a l measurement p r o c e s s f o r Ca, where a e q u a l s 20 ng, and
Q
the d e t e c t i o n l i m i t i s 66 ng.

When d f i s v e r y l a r g e , d i s s i m p l y 2z o r 3.29. F o r fewer

degrees o f freedom, 2 t y i e l d s a c o n s e r v a t i v e e s t i m a t e , b u t a s t i l l
b e t t e r e s t i m a t e d e r i v e s from t h e f o l l o w i n g e x p r e s s i o n (68, p. 252):
d * 3.29 (1 + 0.71/df) (4)
T h i s f o r m u l a i s a c c u r a t e t o about 1% o r b e t t e r f o r d f > 8. F o r 4 -
7 degrees o f freedom t h e c o r r e c t v a l u e s a r e 4.07, 3.87, 3.75, and
3.68. To i l l u s t r a t e , l e t us suppose t h a t 5 p a i r e d y, B
o b s e r v a t i o n s were made and t h e mean d i f f e r e n c e and e s t i m a t e d
s t a n d a r d e r r o r were 1.8 1.2 mV. The c r i t i c a l l e v e l f o r 4 degrees
o f freedom would be t s - (2.13) (1.2) - 2.6 mV, so t h e c o n c l u s i o n
would be "not d e t e c t e d . " The d e t e c t i o n l i m i t would be da Q -
4.07a . o U s i n g s as an e s t i m a t e f o r a, we would e s t i m a t e LQ as
(4.07)(1.2)=4.9 mV.
3.2.5 uncertaintie i detectio limits Th previou
example r a i s e s an e x t r e m e l
i s known w i t h o u t e r r o r
known. This i s i n c o n t r a s t w i t h the c r i t i c a l l e v e l , which can
always be e x p l i c i t l y c a l c u l a t e d from Student's t and t h e e s t i m a t e d
standard error. We c a n , however, d e r i v e a c o n f i d e n c e i n t e r v a l f o r
Lp from t h e bounds f o r a, g i v e n s and d f . F o r n o r m a l l y d i s t r i b u t e d
2
e r r o r s these bounds c a n be d e r i v e d from t h e x d i s t r i b u t i o n . ( s / a 2 2
i s d i s t r i b u t e d as x / d f . ) One f i n d s , f o r example, t h a t a t l e a s t 13
2
r e p l i c a t e s a r e n e c e s s a r y t o o b t a i n s w i t h i n 50% o f t h e t r u e a (90%
confidence l e v e l ) .
For p r a c t i c a l a p p l i c a t i o n o f d e t e c t i o n l i m i t s -- e.g., i n
m e e t i n g a r e s e a r c h o r r e g u l a t o r y r e q u i r e m e n t - - a " s a f e r " procedure
i s t o quote t h e upper l i m i t f o r Lp . This i n e f f e c t casts the
u n c e r t a i n t y onto i n t h a t a s p e c i f i c v a l u e ( r a t h e r t h a n a range)
can be g i v e n f o r t h e d e t e c t i o n l i m i t , b u t w i t h t h e p r o v i s o t h a t
<0.05 ( w i t h 95% c o n f i d e n c e ) . A straightforward, conservative
t r e a t m e n t f o r d e t e c t i o n d e c i s i o n s and d e t e c t i o n l i m i t s when s i s
e s t i m a t e d from r e p l i c a t i o n i s t h u s : t o use = t s (a=0.05) f o r
d e t e c t i o n d e c i s i o n s ; and t o use - 2 t s ( a / s ) (a=0.05, 0<O.O5) f o r
M
detection limits. From t h e b r i e f t a b l e o f t h e r e l e v a n t q u a n t i t i e s

w h i c h f o l l o w s , we see f o r example w i t h n = 10, 1^ - 1.83 s, and
L = 2 ( 1 . 8 3 ) ( 1 . 6 5 ) s - 6.04 s ( t o be compared w i t h 3.29 a f o r d f - .)
D
[Table I I ; from Ref. 28, p. 8 0 ] .
Table I I . Lp E s t i m a t i o n by R e p l i c a t i o n : S t u d e n t ' s - t and (a/s) -

Bounds v s Number o f O b s e r v a t i o n s
No. o f r e p l i c a t e s : 5 10 13 20 120
Student's-t: 2.13 1.83 1.78 1.73 1.66 1.645
a /s:
U L 2.37 1.65 1.51 1.37 1.12 1.000
A second source o f u n c e r t a i n t y i s a s s o c i a t e d w i t h the

q u a n t i t i e s c o m p r i s i n g t h e o v e r a l l c a l i b r a t i o n f a c t o r A, such as
recovery, instrumental detection e f f i c i e n c y , matrix absorption or
s c a t t e r i n g , e t c . I f A i s determined as a random v a r i a b l e each time
x ( c o n c e n t r a t i o n ) i s e s t i m a t e d , t h e n t h e r e i s no problem; i t s
random e r r o r i s a u t o m a t i c a l l y t a k e n i n t o account t h r o u g h e r r o r

p r o p a g a t i o n o r r e p l i c a t i o n when a x i s estimated. I f t h e same

e s t i m a t e f o r t h e c a l i b r a t i o n f a c t o r i s r e p e a t e d l y used, i t s random
e r r o r has become a b i a s , and t h e bounds ( c o n f i d e n c e i n t e r v a l ) f o r
t h i s b i a s combined w i t h o t h e r p o s s i b l e s o u r c e s o f A - b i a s produce an
uncertainty i n t e r v a l i n the concentration detection l i m i t . The
recommended approach, a g a i n t a k i n g i n t o account t h e p r a c t i c a l
applications f o r detection l i m i t s , i s to t r a n s f e r the uncertainty
to p by t a k i n g t h e upper l i m i t , Sp/A^, as t h e c o n c e n t r a t i o n
d e t e c t i o n l i m i t w i t h p < 0.05.
3.2.6 u n c e r t a i n t y bounds ( s y s t e m a t i c e r r o r ) f o r t h e b l a n k .
I f the p o s s i b i l i t y o f s i g n i f i c a n t bias i n the estimated value f o r
the b l a n k i s n o t t a k e n i n t o a c c o u n t , t h e r e s u l t a n t d e t e c t i o n
d e c i s i o n s and l i m i t s may be much t o o o p t i m i s t i c . An upper l i m i t
f o r t h i s b i a s component c a n be i n c o r p o r a t e d i n t o S and S estima-
c D
t i o n j u s t as i t i s i n t o t a l u n c e r t a i n t y i n t e r v a l e s t i m a t i o n , by
e x t e n d i n g t h e random u n c e r t a i n t y ( c o n f i d e n c e ) l i m i t by t h e upper
bound f o r b i a s . Thus
2 S '. c The d e t e c t i o
r a t i o n a l e f o r t h i s p r o c e d u r e i s i n d i c a t e d i n F i g . 9. F o r a number
o f measurement d i s c i p l i n e s , e x p e r i e n c e d i c t a t e s r e a s o n a b l e v a l u e s
f o r r e l a t i v e l i m i t s f o r b l a n k and c a l i b r a t i o n f a c t o r b i a s [ 4 B > 4 A ] -
D e f a u l t v a l u e s o f 5% and 10%, r e s p e c t i v e l y , have been s u g g e s t e d
(28) and t e n t a t i v e l y c o n f i r m e d (57) f o r r a d i o a c t i v i t y m o n i t o r i n g ,
f o r example. I n t h i s c a s e , S ' - 1.645 o + 0 . 0 5 B = S + 0 . 0 5 B,
c Q c
and x ' - 1.1 (2 S ' ) / A - 1.1 x + 0.11 BEA, where x = 3.29 o /k

D c D D Q
and BEA i s t h e background e q u i v a l e n t a c t i v i t y . Thus f o r p a i r e d

measurements w i t h B - 500 counts ( a - J2B - 31.6 c o u n t s ) , and A =
Q
5.0 c o u n t / p C i , S ' - 52 + 25 - 77 c o u n t s , and x ' - 22.9 + 11.0 =

c D
33.9 p C i . C l e a r l y , d e t e c t i o n i n t h i s case i s n e i t h e r fully

s t a t i s t i c a l nor f u l l y n o n - s t a t i s t i c a l . Balancing the l i m i t s
imposed by t h e s t a t i s t i c s o f s i g n a l d e t e c t i o n w i t h those d e r i v e d
from o u r knowledge ( o r i g n o r a n c e ) o f t h e measurement p r o c e s s i s
e s s e n t i a l f o r m e a n i n g f u l d e c i s i o n making. H i s t o r i c a l use o f a
m u l t i p l e o f t h e b l a n k i s perhaps more r e a d i l y u n d e r s t o o d a l s o ,
t h r o u g h t h e f o r m a l i n c o r p o r a t i o n o f t h e term 0.11 BEA.
3.2.7 o p t i m i z a t i o n and i t e r a t i o n : figure of merit. Optimal
d e t e c t i o n l i m i t s a r e sometimes t r e a t e d t h r o u g h t h e m a x i m i z a t i o n o f
some s o r t o f f i g u r e o f m e r i t (FOM) such as S//B, e t c . S i m p l i s t i c
FOM's t e n d t o i g n o r e complex dependence o f Lp ' s on measurement
c o n d i t i o n s , s y s t e m a t i c e r r o r components, and t h e e x p l i c i t n a t u r e o f
the sample. As shown i n (35) f o r example t h e v a r i a t i o n o f r a d i o -
1
a c t i v i t y d e t e c t i o n l i m i t s w i t h c o u n t i n g time may range from t " t o
+ 1
t . S i n c e d e t e c t i o n l i m i t s may be sample-dependent, because o f
i n t e r f e r e n c e and m a t r i x e f f e c t s , i t e r a t i v e estimation o f the
d e t e c t i o n l i m i t i s sometimes r e q u i r e d . Changes i n t h e measurement
p r o c e s s may a l s o be n e c e s s a r y i f such sample dependence f o r c e s t h e
a c t u a l d e t e c t i o n l i m i t above t h e c o r r e s p o n d i n g r e g u l a t o r y l i m i t .
3.2.8 multicomponent d e t e c t i o n l i m i t s . When one l e a v e s t h e
r e a l m o f " s i m p l e " y - B n e t s i g n a l e s t i m a t i o n , m o d e l i n g and l i n e a r
or n o n - l i n e a r l e a s t squares computations a r e g e n e r a l l y r e q u i r e d f o r
component e s t i m a t i o n . F o r t h e l i n e a r multicomponent model i t i s
p o s s i b l e t o e s t i m a t e t h e d e t e c t i o n l i m i t as a c l o s e d e x p r e s s i o n ,
p r o v i d e d t h a t a l l i n t e r f e r i n g a n a l y t e s a r e i n c l u d e d and t h e e r r o r s
(variance), constant. Weighted least squares calculations
i n v o l v i n g Poisson or other concentration-dependent statistical

F i g . 9. E f f e c t o f B i a s on D e t e c t i o n L i m i t s . A l l o w a n c e for
bounds f o r b i a s [ A ] i n c r e a s e s the c r i t i c a l l e v e l by A , and
M M the
d e t e c t i o n l i m i t by t w i c e t h a t amount ( s i m p l e d e t e c t i o n ) , t a k i n g the
s i g n f o r the uncompensated b i a s as unknown. a and /3 are now
inequalities i e , a, < 0.05.

w e i g h t s r e q u i r e an i t e r a t i v e Lj, e s t i m a t e (13.61). A relatively

s i m p l e i n e q u a l i t y e x p r e s s i o n f o r Lp can be g i v e n , however, by u s i n g
the r e s u l t s o f a l i n e a r model f i t , where the component o f i n t e r e s t
i s p r e s e n t a t o r above i t s d e t e c t i o n l i m i t ( 2 8 ) . That i s , x D <
3.29 xa where a x i s the s t a n d a r d e r r o r f o r the l e a s t squares
e s t i m a t e o f the component i n q u e s t i o n . The b a s i s f o r t h i s
i n e q u a l i t y i s t h a t the s t a n d a r d e r r o r f o r any s p e c i f i c component
w i l l be a p p r o x i m a t e l y c o n s t a n t o r i n c r e a s e w i t h i t s i n c r e a s i n g
concentration i n a mixture, a l l e l s e remaining equal. Special
multicomponent problems a s s o c i a t e d w i t h s e l e c t e d t y p e s o f p a t t e r n
m a t c h i n g and m u l t i p l e independent d e c i s i o n s , as i n the d e t e c t i o n o f
s e v e r a l i s o l a t e d s p e c t r a l o r chromatographic peaks, w i l l be t r e a t e d
i n the n e x t s e c t i o n .
3.2.9 e f f e c t s o f g r a d u a l l y changing d i s t r i b u t i o n f u n c t i o n s
and/or a's. For a - n o r m a l l y d i s t r i b u t e d e r r o r s w i t h con-
s t a n t , known v a r i a n c e d simpl p a i r e d estimate ( B)
Ljj/Lg-2. This simple r a t i
o f t h e s e c o n d i t i o n s ar
o l d s (a 1) has a l r e a d y been noted. I n the example d i s c u s s e d i n
the l a s t p a r a g r a p h o f the p r e c e d i n g s e c t i o n , a was v a n i s h i n g l y
s m a l l when was s e t t o 3 4 a . Q The r a t i o L p / I ^ i n t h i s case was
1.09 (61). S i m i l a r l y , the r a t i o i s u n i t y when the c o n c e p t u a l
d i f f e r e n c e between and Lp i s o v e r l o o k e d , such t h a t 0 - 0.50.
Changing c d f ' s i s another m a t t e r . Because the overall
d e t e c t i o n p r o c e s s i n e f f e c t r e l a t e s t o the d i s c r i m i n a t i o n o f n e t
s i g n a l s a t the d e t e c t i o n l i m i t from n u l l s i g n a l s , one i s f a c e d w i t h
the p o s s i b i l i t y o f two d i f f e r e n t d i s t r i b u t i o n s a t the two l e v e l s ,
S-0 and S-S . D T h i s problem does n o t a r i s e i n making d e t e c t i o n
d e c i s i o n s , hqwever, f o r S c depends o n l y on a , the Q standard
d e v i a t i o n o f S when i t s t r u e (mean) v a l u e i s z e r o . Two cases w i l l
be c o n s i d e r e d , a) the P o i s s o n c o u n t i n g d i s t r i b u t i o n , w h i c h changes
shape and ( r e l a t i v e ) d i s c r e t e n e s s w i t h i n c r e a s i n g s i g n a l l e v e l ; and
b) the normal d i s t r i b u t i o n , where a i n c r e a s e s w i t h c o n c e n t r a t i o n , a
common o c c u r r e n c e i n a n a l y t i c a l c h e m i s t r y . A t h i r d case o f some
importance f o r e n v i r o n m e n t a l measurements i s the d i s t r i b u t i o n a l
p e r t u r b a t i o n w h i c h o c c u r s as one adds normal measurement e r r o r s t o
log-normal blank v a r i a t i o n s .
The P o i s s o n d i s t r i b u t i o n i s d e c i d e l y asymmetric and d i s c r e t e
( i n a r e l a t i v e sense) a t the l o w e s t l e v e l s . I n f a c t , when the
e x p e c t e d (mean) v a l u e o f the P o i s s o n parameter - - i n t h i s c a s e , the
b l a n k - - i s s m a l l e r t h a n 0.05 c o u n t s , the c r i t i c a l l e v e l ( y , g r o s s c
counts) equals zero. ( T h i s q u a n t i t y i s n e c e s s a r i l y an i n t e g e r f o r

the P o i s s o n d i s t r i b u t i o n , and a must be t r e a t e d as an i n e q u a l i t y
[o<0.05].) The d e t e c t i o n l i m i t ( y , g r o s s c o u n t s ) t h e n e q u a l s 3.00
D
( n o t n e c e s s a r i l y an i n t e g e r ) , so y / y D i n t h i s case i s i n f i n i t e .
c
W i t h i n c r e a s i n g s i g n a l l e v e l ( c o u n t s ) the P o i s s o n d i s t r i b u t i o n
approaches n o r m a l i t y , so "the u s u a l e q u a t i o n s a p p l y , " and IÂc
approaches 2.0. For B - 1.0, I ^ / I ^ 3.4; f o r B>5, 1^ 2.7+2-1^,
w i t h Lg - 1.645 76 i s a good a p p r o x i m a t i o n . See (28) f o r a more
e x t e n s i v e t r e a t m e n t o f extreme, l o w - l e v e l c o u n t i n g s t a t i s t i c s .
P o i s s o n a's i n c r e a s e w i t h 7(S+B). A l i n e a r i n c r e a s e o f a w i t h
c o n c e n t r a t i o n [ a ( y ) - a + mS] , however, i s common f o r many a n a l y -
B
t i c a l methods. S i n c e S - z a and S - S + z a , the i n c r e a s e i n a

c Q D c D
i n p a s s i n g from S=0 t o S-S means t h a t S > 2 S .

D D c A c l o s e d expres-
s i o n , however, may be d e r i v e d u s i n g the above l i n e a r model. (To

simplify the notation i n the remainder of t h i s paragraph, S and S c D
w i l l be represented as C and D respectively.) Let us consider

f i r s t the case where a precise estimate fi i s a v a i l a b l e f o r B.
Standard deviations are given by:
a, assumed v a r i a t i o n :
- a(S+B) - a + mS a y B
a, n u l l signal [S - 0]: - a(B-fi) a a Q B
a, detection l i m i t [S - D] :
- a(D+B - fi) a(D+B) - a D
a + mD B
For normal random errors and a - 0 - 0.05, the c r i t i c a l l e v e l

and detection l i m i t are defined as:
C - za Q ~ za fi - 1.645 a B
D - C + za D C + z(a B + mD)
The l a s t equation
D - 2C/(l-zm)
For a--0.05, i . e . , z-1.645, an important conclusion emerges: the
detection l i m i t does not exist for m > 1/z - 0.61. This may be
academic, however, since so large a slope i s u n l i k e l y f o r any
reasonable a n a l y t i c a l method. A slope of 10%, however, would
r e s u l t i n D/C - 2.39. To i l l u s t r a t e , l e t us take the blank
standard deviation for the measurement of toluene i n a i r , by a
f u l l y s p e c i f i e d method of sampling and gas chromatographic analy-
s i s , to be 0.21 Mg/L. The c r i t i c a l l e v e l f o r detection decisions,
assuming normality, would then be 1.645 (0.21) - 0.34 /xg/L. The
corresponding detection l i m i t would be 2.39(0.34 ) - 0.83 /ig/L- 5
For the general case, where B i s estimated from n r e p l i c a t e s ,

the algebra i s only s l i g h t l y more complicated. The variance of the
estimated net signal i s now given by:
A M S 2 + A 2 N M S 2 ( 7 + M S
v
s " <B + ) B / " V
o +
< B )
thus, defining y-(n+l)/n:
a
V
o - l - l 1 a n d
V D - ag - V D + mD (2a B + mD)
From these relations and the d e f i n i t i o n s f o r C [C - za ] Q and

D[D-C+za ], i t i s r e l a t i v e l y straightforward to show that:
D
2
D - 2C [1 + zm/7^]/[l - (zm) ] and a / a D Q - D/C - 1
Thus for ?y-2 (paired comparison), and m and z as before (0.1,

1.645), D/C - 2.29 and a^/a - 1.29. The asymptotic r e s u l t (D-2C)
Q
follows of course when the slope m i s n e g l i g i b l e .

3.2.10 black boxes and hidden algorithms. With the advent of
"user f r i e n d l y " (and proprietary) software and automated data
reduction and even automated instrument systems which y i e l d f i n a l
r e s u l t s only, a cautionary note must be sounded. That i s , when the
computational scheme i s not f u l l y and e x p l i c i t l y described, and
when the software i s not exhaustively studied and tested, erroneous
r e s u l t s may emerge. Worse s t i l l , there may be no way of
recognizing such results as erroneous, particularly i f the

i n s t r u m e n t a l system i s d e s i g n e d i n such a manner t h a t the raw

e x p e r i m e n t a l d a t a cannot be r e t r i e v e d f o r a l t e r n a t i v e methods o f
computation. I t i s i n a p p r o p r i a t e i n t h i s c h a p t e r t o document the
problems a r i s i n g , b u t i t may be h e l p f u l to g l i m p s e a t t h e i r n a t u r e .
I n e p t o r u n l u c k y programming and i n a c c u r a t e s t o r e d parameters w i l l
always cause d i f f i c u l t i e s , b u t t h i s o f c o u r s e i s not r e s t r i c t e d to
the domain o f l o w - l e v e l measurement. Problems o f s p e c i a l c o n c e r n
f o r r e l i a b l e and e f f i c i e n t d e t e c t i o n w h i c h have come to the
a t t e n t i o n of t h i s author i n c l u d e : a) T h r e s h o l d s w h i c h are
a u t o m a t i c a l l y s e t so h i g h t h a t d e t e c t i o n power i s s e r i o u s l y eroded;
b) A l g o r i t h m s (and component models) w h i c h are d a t a dependent.
T h i s i s e s p e c i a l l y a problem when peaks are m a r g i n a l l y d i s c e r n a b l e ,
w i t h peak e s t i m a t i o n a l g o r i t h m s s w i t c h i n g r u l e s depending on the
magnitude o r apparent p r e s e n c e o f a peak; c) e s t i m a t i o n and s e a r c h
r o u t i n e s based on inadequate models or i n a d e q u a t e l y a c c o u n t i n g f o r
the e f f e c t s o f n o n - l i n e a r e s t i m a t i o n ; d) d e c i s i o n or d e t e c t i o n
a l g o r i t h m s f o r w h i c h assumption
(and possibly incorrect)
e s p e c i a l l y when peaks are not found, and the consequent i n a b i l i t y
t o i n v e s t i g a t e e x t r a s o u r c e s o f v a r i a b i l i t y or e r r o r s i n assump-
t i o n s , models or d a t a .
Peak s e a r c h or "model s e a r c h " and more g e n e r a l l y o p t i m i z a t i o n
r o u t i n e s t h a t are sometimes h e u r i s t i c and o p e r a t e s t r i c t l y i n an
e m p i r i c a l f a s h i o n on the d a t a a t hand, deserve a n o t h e r comment.
That i s , a t the l o w e s t l e v e l s and e s p e c i a l l y a t the S - 0 extreme
[ n u l l h y p o t h e s i s ] such g e n e r a l l y n o n - l i n e a r r o u t i n e s may p r o v i d e no
S e s t i m a t e s , e s p e c i a l l y when n e g a t i v e , thus p r o d u c i n g b i a s e d and
skewed S d i s t r i b u t i o n s . Once a peak o r model i s a u t o m a t i c a l l y
chosen from the n o i s y d a t a , the a l g o r i t h m s w i t c h e s , f r e q u e n t l y t o a
l i n e a r estimation algorithm. The problem i s t h a t the s w i t c h p o i n t
v a r i e s , b e i n g n o i s e c o n t r o l l e d ; a l s o the e s t i m a t i o n algorithm
seldom g e t s t o o p e r a t e when the n u l l h y p o t h e s i s [S=0] i s t r u e . a
Q
i s not o b t a i n e d , and the normal d i s t r i b u t i o n h y p o t h e s i s t e s t i n g

a p p a r a t u s cannot be a p p l i e d a t the l o w e s t s i g n a l l e v e l s . Perhaps
t h i s i s why the i n t e r n a t i o n a l gamma r a y peak d e t e c t i o n e x e r c i s e
o r g a n i z e d by the I n t e r n a t i o n a l Atomic Energy Agency found the
" v i s u a l " method o f peak d e t e c t i o n more s u c c e s s f u l t h a n a l l o t h e r s ,
i n c l u d i n g the most s o p h i s t i c a t e d computer based schemes (.58) .
The "model s e a r c h " i s s u e i s more p r o f o u n d . I n multicomponent
c h e m i c a l a n a l y s i s , o p t i m a l models f o r e s t i m a t i o n (number and n a t u r e
o f components) are o f t e n chosen a u t o m a t i c a l l y and e m p i r i c a l l y , f o r
example by a p p l y i n g i t e r a t i v e , n o n - l i n e a r o p t i m i z a t i o n r o u t i n e s ,
and q u i t e f r e q u e n t l y n o n - n e g a t i v i t y c o n s t r a i n t s . Such a u t o m a t i c
chemical model b u i l d i n g , a c c o m p l i s h e d by suppressing (often
l e g i t i m a t e ) n e g a t i v e e s t i m a t e s , d e s e r v e s c a r e f u l s c r u t i n y . I t may
be even more m i s l e a d i n g than zero s u p p r e s s i o n w i t h s i m p l e measure-
ments, e s p e c i a l l y when n o i s e and m u l t i c o l l i n e a r i t y are l a r g e .
I l l u s t r a t i o n s o f some o f these l i m i t a t i o n s , w h i c h are unique
f o r l o w - l e v e l d a t a and t h e r e f o r e m e a n i n g f u l d e t e c t i o n l i m i t s , may
be found i n r e f e r e n c e s 35, 57, and 70. F i g . 2 i n the f i r s t r e f e r -
ence i l l u s t r a t e s an extreme, y e t not uncommon problem; q u i t e
v i s i b l e s p e c t r a l peaks have f a i l e d t o be d e t e c t e d by the s o f t w a r e .
3.2.11 quality. One s o l u t i o n f o r inadequate or i n c o r r e c t
approaches t o d e t e c t i o n -- i n c l u d i n g c o n t r o l o f b o t h f a l s e p o s i -
t i v e s and f a l s e n e g a t i v e s - - i s the i n c o r p o r a t i o n o f known and

b l i n d s t a n d a r d r e f e r e n c e samples and r e f e r e n c e s i m u l a t e d data.

Such means f o r c o n t r o l are w e l l e s t a b l i s h e d f o r t r a c e a n a l y s i s , b u t
t h e y have r a r e l y been b r o u g h t t o b e a r on the d e t e c t i o n problem.
I n t e r l a b o r a t o r y l o w - l e v e l t e s t d a t a , though q u i t e r a r e , have p r o v e n
most i n f o r m a t i v e (58.70). D i r e c t v a l i d a t i o n and c o n t r o l o f a and p
e r r o r s s h o u l d be made r o u t i n e l y w i t h b l i n d i n t e r l a b o r a t o r y samples
and/or d a t a r e p r e s e n t a t i v e o f b l a n k s and samples a t o r near
d e t e c t i o n (or r e g u l a t o r y ) l i m i t s , r e s p e c t i v e l y . E v a l u a t i o n of sets
o f r e s u l t s v i a ROC c u r v e s c o u l d , i n t u r n , be q u i t e f r u i t f u l , f o r
the q u a l i t y o f the l o w - l e v e l measurements would be r e f l e c t e d i n
the l o c i o f the ROC curves, independent of the particular decision
rules employed.
3.3 Discrimination Limits. Multiple Detection Decisions, and

Patterns. When the n u l l h y p o t h e s i s i s d e f i n e d as z e r o a n a l y t e
c o n c e n t r a t i o n beyond th b l a n k s i g n a l abov th baselin
o r background, i t i s a p p r o p r i a t
d e t e c t i o n process. I
i n t e r e s t i n g t o c o n s i d e r the a b i l i t y t o d i s c r i m i n a t e c o n c e n t r a t i o n s
from a f i x e d non-zero r e f e r e n c e l e v e l , o r d i s c r i m i n a t e p a t t e r n s o r
"chemical f i n g e r p r i n t s " from a r e f e r e n c e pattern. Multiple
h y p o t h e s i s t e s t i n g d e c i s i o n s form a n a t u r a l l i n k between t h e s e two
t y p e s o f d i s c r i m i n a t i o n , and i t becomes c l e a r t h a t b o t h f i x e d l e v e l
" r e c o g n i t i o n " and c h e m i c a l p a t t e r n r e c o g n i t i o n f a l l under the same
s t a t i s t i c a l frame-work as z e r o l e v e l a n a l y t e d e t e c t i o n . Both
a s p e c t s o f Q u a l i t a t i v e A n a l y s i s ( d e t e c t i o n , i d e n t i f i c a t i o n ) share
the same p r o b a b i l i s t i c f o u n d a t i o n , i n c l u d i n g h y p o t h e s i s s p e c i f i c a -
t i o n [H , H ] , d e c i s i o n c r i t e r i a , and type I [a] and type I I [ft]
Q A
e r r o r s [Ref. 8; pp. 233, 239]. In a l l cases, i t i s extremely

important to recognize t h a t the r e s p e c t i v e d i s c r i m i n a t i o n o r
d e t e c t i o n l i m i t s c h a r a c t e r i z e the measurement p r o c e s s , not a
particular result. (As always, r e s u l t s a r e t e s t e d by comparison t o
the c o r r e s p o n d i n g c r i t i c a l l e v e l . ) Our o b j e c t i v e i s t o e v a l u a t e
the intrinsic capabilities of CMP's, often shaping these
c a p a b i l i t i e s t o meet s p e c i f i c p r a c t i c a l o r r e s e a r c h needs.
3.3.1 lower and upper r e g u l a t o r y l i m i t s : b a l a n c i n g r i s k s and
costs. We have n o t e d t h a t d e t e c t i o n l i m i t s d i c t a t e d by r e g u l a t o r y
concerns have been surrounded by c o n s i d e r a b l e c o n f u s i o n , d i s c r e p a n t
s t a t i s t i c a l and ad hoc formulations, ignorance, and even m i l d
deception. The apparent d e c e p t i o n i s r e l a t e d t o the l a c k o f
general understanding o r agreement c o n c e r n i n g the appropriate
n a t u r e and magnitude o f the e r r o r o f the second k i n d (/3, f a l s e
negative). By i g n o r i n g i t s p r e s e n c e , whether i n t e n t i o n a l o r n o t ,
those who must meet r e g u l a t o r y demands g e n e r a t e a fi/a imbalance
where, a t 50%, f a l s e n e g a t i v e s may exceed f a l s e p o s i t i v e s by n e a r l y
a f a c t o r o f 400. One j u s t i f i c a t i o n i s t h a t i d e n t i c a l l y z e r o con-
c e n t r a t i o n s cannot e x i s t anyway, and v e r y s m a l l concentrations
cannot be e f f e c t i v e l y d i s t i n g u i s h e d from the b l a n k . A related,
v e r y i m p o r t a n t o b s e r v a t i o n i s t h a t s m a l l non-zero c o n c e n t r a t i o n s
w i l l be " d e t e c t e d " on o c c a s i o n , n e c e s s a r i l y more f r e q u e n t l y t h a n
the f a l s e p o s i t i v e c o n s t r a i n t [a] p l a c e d on the b l a n k . To reduce
the p e n a l t y w h i c h might be a s s o c i a t e d w i t h the o c c a s i o n a l d e t e c t i o n
o f such s m a l l c o n c e n t r a t i o n s , i t i s o f c o u r s e h e l p f u l t o reduce a
f o r the b l a n k s t i l l f u r t h e r but t h i s s h o u l d be done o p e n l y , not
by s u b t e r f u g e .

To meet such legitimate concerns, while at the same time

keeping an open, r e a l i s t i c , and balanced view of f a l s e p o s i t i v e s
and negatives, we recommend the s u b s t i t u t i o n of lower and upper
regulatory l i m i t s whose difference i s the discrimination l i m i t
(AJJ) i n place of the n u l l l i m i t [zero] and the single, analyte
detection l i m i t . To i l l u s t r a t e the suggested approach, F i g . 2 has
been modified i n F i g . 10 to indicate a non-zero lower l i m i t ( L Q ) ,
and an upper l i m i t (Lp). AS before, the upper l i m i t i n the
s o c i e t a l or regulatory s e t t i n g would be established at such a l e v e l
that the concentration or event of concern would be r e l i a b l y
detected [/3-0.05] when i t s net "cost" to society crossed the l i m i t
of a c c e p t a b i l i t y . The lower l i m i t from which the upper l i m i t must
be r e l i a b l y discriminated, i s new: i t s l e v e l i s established such
that the penalty for "detecting" a very small concentration i s
likewise acceptable. Such penalties can be quite r e a l , e s p e c i a l l y
i n terms of intangibles such as public alarm (21) or i n d i r e c t
long-term negative perception
regulatee. In F i g . 1
context of earthquake detection, with the a i d of hypothetical
p o s i t i v e and negative cost d i f f e r e n t i a l s which would define the
"trigger points" for and Lp.
I t i s important that the (regulatory) l e v e l - s e t t i n g process
for these l i m i t s be decoupled from their estimation from the
c h a r a c t e r i s t i c s of the measurement process. The former i s a
s o c i o p o l i t i c a l matter involving complex r i s k assessment issues ( 4 ) ,
whilst the l a t t e r l i e s i n the domain of the s c i e n t i s t . The
s c i e n t i f i c r e s p o n s i b i l i t y i s met once the discriminable l i m i t s l i e
within those desired by society. Note that the discrimination
l i m i t Ap i s here defined as the difference Lp -Lp such that a, g
each equal 0.05. I t i s i n t e r e s t i n g next to consider p r e c i s i o n
requirements, e.g., at the upper l i m i t , as compared to those for
the conventional detection l i m i t . Taking Lp to be 50% of Lp , the
r e l a t i v e standard deviation at this Lp would be about 15%, i n
contrast with 30% for the conventional detection l i m i t . (The
change i s e n t i r e l y due to the introduction of the non-zero Lp ; the
magnitude of Lp i s unchanged.) The p r e c i s i o n (RSD at Lp) would be
"quantitative" ( 1 0 % ) , once Lp equals 2/3 of Lp . Quite possibly the
(subconscious) need f o r such discrimination c a p a b i l i t y i s the
underlying motivation of those who c a l l f o r abandoning detection
limits and hypothesis testing in favor of "quantitative"
measurements.
The discrimination l i m i t as depicted i n F i g . 10 has two other
important applications, one i n business and one i n science. In
business matters involving trade or regulation, one may face the
task of "proving" the product or waste stream l e v e l exceeds or does
not exceed some prescribed value, such as the (upper) regulatory
limit. Because of measurement error, the a b i l i t y to accomplish
t h i s i s limited, and i n fact i t i s set by the size of the dis-
crimination l i m i t . Balancing of costs w i l l again generally f i x the
magnitude of A . D Penalties w i l l l i k e l y increase with greater
apparent departures from s p e c i f i c a t i o n s ; and the a b i l i t y to defend
departures as small, or attack departures as large depends upon the
producer's or consumer's discrimination l i m i t . The discrimination
l i m i t , hence p r e c i s i o n of analysis, can only be improved with
increased a n a l y t i c a l costs. In the socioeconomic arena decision

Fig. 10. Discrimination Limits. Curve F" represents the loss

to society as a function of earthquake magnitude; F* represents the
cost of avoidance (evacuation, etc.), the dashed portion simulating
indirect costs associated with f a l s e alarms -- eg, mental anguish,
damaged c r e d i b i l i t y , lawsuits, etc. Points of imbalance between F*
and F" which exceed what i s acceptable to society are taken as
lower and upper regulatory l i m i t s , which must be matched by
corresponding lower ( L ) and upper (Lp) measurement l i m i t s whose
Q
difference i s the Discrimination Limit ( A ) . D A non-zero lower

l i m i t forces an improved p r e c i s i o n requirement i n comparison to the
"simple" Lp of F i g . 2.

t h e o r y may be h e l p f u l f o r d e r i v i n g the a p p r o p r i a t e b a l a n c e between

p e n a l t i e s and a n a l y t i c a l costs ( t h e r e f o r e the r e q u i s i t e A ) , D
p a r t i c u l a r l y t a k i n g i n t o c o n s i d e r a t i o n w h i c h p a r t y has the burden

of proof. (Note t h a t the " c o s t " differential a s s o c i a t e d w i t h the
burden of proof i s e q u i v a l e n t t o the s i z e o f the [measurement] dead
zone around the r e g u l a t o r y o r s p e c i f i c a t i o n l i m i t -- i . e . , the sum
o f the "producer's" and "consumer's" d i s c r i m i n a t i o n l i m i t s . )
The s c i e n t i f i c a p p l i c a t i o n i n v o l v e s " i d e n t i f i c a t i o n " i n i t s
s i m p l e s t sense. That i s , i f Lp and Lp are t r e a t e d as unique o r
i d e n t i f y i n g concentrations or isotope r a t i o s , or c h a r a c t e r i s t i c
e n e r g i e s o r w a v e l e n g t h s , e t c . , t h e n the measurement p r o c e s s must be
d e s i g n e d so t h a t A D i s s u f f i c i e n t l y s m a l l t o d i s t i n g u i s h between
t h e s e two c l a s s e s . I n analogy w i t h the d e t e c t i o n power (1-/0
c h a r a c t e r i z i n g the d e t e c t i o n l i m i t ( g i v e n a ) , one f i n d s the power
a s s o c i a t e d w i t h A d e s c r i b e d as " d i s c r i m i n a t o r y power" (12, p. 517)
D
o r " r e s o l v i n g power." T h i s u n i v a r i a t e s t a t i s t i c a l approach t o

i d e n t i f i c a t i o n shares muc
OC and ROC c u r v e s ar
p o s i t i v e s and n e g a t i v e s , and f o r comparing c a p a b i l i t i e s o f a l t e r n a -
t i v e measurement (and c o m p u t a t i o n a l ) t e c h n i q u e s . I n a d d i t i o n , the
d i f f e r e n c e between d e s i g n o f the measurement p r o c e s s t o a c h i e v e a
g i v e n d e t e c t i o n o r i d e n t i f i c a t i o n c a p a b i l i t y and outcome ( s p e c i f i c
r e s u l t ) i s s t i l l m a n i f e s t i n an u n c e r t a i n r e g i o n -- i . e . , r e s u l t s
f a l l i n g w i t h i n the RUD[region o f u n c e r t a i n d e t e c t i o n ] o r RUI [ r e g i o n
of uncertain identification] may be detected or identified,
r e s p e c t i v e l y , by chance b u t t h i s cannot be " a s s u r e d " ( a , 0 - 0.05)
a priori. (See " m u l t i c h a n n e l i d e n t i f i c a t i o n , below, f o r f u r t h e r
discussion.)
3.3.2 impurity detection. A s p e c i a l issue involving dis-
crimination limits in analytical chemistry, having broad
i m p o r t a n c e , i s the d e t e c t i o n o f i m p u r i t i e s o r contamination.
C o n c e p t u a l l y , t h i s can be t r e a t e d as a d i r e c t outgrowth o f the
" i d e n t i f i c a t i o n " or d i s c r i m i n a t i o n of s i n g u l a r c l a s s e s charac-
t e r i z e d by unique v a l u e s o f a c o n t i n u o u s v a r i a b l e , as d e s c r i b e d i n
the p r e c e d i n g p a r a g r a p h . I n F i g . 11 c l a s s - 0 and c l a s s - A are shown
a t s e p a r a t e unique ( i d e n t i f y i n g ) l o c a t i o n s o f a c o n t i n u o u s measure-
ment v a r i a b l e x . As d e p i c t e d , the s e p a r a t i o n o f t h e s e two c l a s s e s
i
f a r exceeds the d i s c r i m i n a t i o n l i m i t A , so i d e n t i f i c a t i o n o f a
D
pure component ( i n t h i s 2-component u n i v e r s e ) w i l l p r e s e n t no

problem. I f component-0 i s c o n t a m i n a t e d by a s m a l l admixture o f
component-A, however, t h e r e e x i s t s a l i m i t [A ] below w h i c h a D
c o n t a m i n a t e d sample w i l l be i n d i s t i n g u i s h a b l e from the pure

component-0. The minimum d e t e c t a b l e c o n t a m i n a t i o n i s n u m e r i c a l l y
e q u a l t o A , when A i s e x p r e s s e d r e l a t i v e t o the c l a s s s e p a r a t i o n
D D
( x - XQ) -- i . e . , as a mole f r a c t i o n o r m i x i n g r a t i o . (Note t h a t

A
" m i x i n g " can o c c u r as p h y s i c a l m i x i n g o f m i s c i b l e c h e m i c a l s p e c i e s ,

o r i t can a r i s e from s u p e r p o s i t i o n o f s i g n a l s from d i f f e r e n t
s o u r c e s w i t h i n the same d e t e c t o r . )
Two fundamental observations follow. First, class
s e p a r a b i l i t y and i m p u r i t y d e t e c t i o n power degrade w i t h i n c r e a s i n g
v a r i a n c e o f the x A d i s t r i b u t i o n s , w h i c h i n t u r n , depends on the
measurement p r e c i s i o n and t h e r e f o r e the d e t e c t i o n l i m i t s f o r the
two components. T h i s d i r e c t , and q u a n t i f i a b l e c o u p l i n g between
pure component d e t e c t i o n l i m i t s , component i d e n t i f i c a t i o n and
r e s o l u t i o n , and i m p u r i t y d e t e c t i o n i s most i m p o r t a n t , though

scarcely surprising. Second, i f one has s u f f i c i e n t knowledge o f

the " c h e m i c a l u n i v e r s e " i . e . , -x. l o c a t i o n s f o r t h e e n t i r e

p o p u l a t i o n o f H 's -- t h e n f o r any H o f i n t e r e s t , one c a n deduce

A 0
the maximum s y s t e m a t i c e r r o r due t o u n d e t e c t e d c o n t a m i n a t i o n by

estimating A D f o r the " c l o s e s t " impurity source. I f this dis-
c r i m i n a t i o n l i m i t i s u n a c c e p t a b l e , r e d e s i g n o f t h e CMP i s i n o r d e r .
R e l i a b l e estimation o f systematic e r r o r bounds d e r i v i n g from
undetectable contamination, o r u n d e t e c t a b l e model e r r o r i s one o f
the i m p o r t a n t needs f o r a c c u r a t e a n a l y t i c a l r e s u l t s . Thoughtful
c o n s i d e r a t i o n o f t h e c o u p l i n g between e x p e r i m e n t a l d e s i g n , and
component d e t e c t i o n and d i s c r i m i n a t i o n l i m i t s , s u p p o r t e d by
e x c e l l e n t s c i e n t i f i c knowledge c o n c e r n i n g t h e H u n i v e r s e o f f e r s
A
one o f t h e most r e l i a b l e and o b j e c t i v e s o l u t i o n s t o t h i s problem.

An a s t u t e e x a m i n a t i o n o f these i s s u e s , e m p h a s i z i n g t h e u n i v e r s e o f
p o t e n t i a l contaminants has been p r o v i d e d by Rogers (6) . F o r
s i m p l i c i t y , t h e d i s c r i m i n a t i o proble presented her i
d i m e n s i o n (one measuremen
discrimination a r e obviou
i n c r e a s e d d e t e c t i o n and d i s c r i m i n a t i o n power, as one compares o r
"matches" unique m u l t i v a r i a b l e p a t t e r n s i n p l a c e o f c h a r a c t e r i s t i c
v a l u e s o f a s i n g l e v a r i a b l e . (See below.)
3.3.3 multiple detection decisions. I f a number o f
d e t e c t i o n o r d i s c r i m i n a t i o n d e c i s i o n s a r e made i n t h e c o u r s e o f a
measurement, t h e o v e r a l l p r o b a b i l i t i e s o f f a l s e p o s i t i v e s and f a l s e
n e g a t i v e s a r e a c c o r d i n g l y a l t e r e d . We c o n s i d e r two c a s e s : f i r s t ,
where t h e i n d i v i d u a l t e s t s a r e u n r e l a t e d o r " s e r i a l " , and second,
where " p a r a l l e l " t e s t s a r e made, as i n p a t t e r n recognition.
Independent, s e r i a l t e s t s c h a r a c t e r i z e t h e d e t e c t i o n o f i s o l a t e d
s p e c t r a l peaks, as i n m u l t i c h a n n e l gamma r a y s p e c t r o s c o p y , as w e l l
as r e s i d u a l s f o l l o w i n g d a t a a n a l y s i s , and even r e p l i c a t i o n e x p e r i -
ments and c o n t r o l c h a r t s . I n a l l o f these cases, the o v e r a l l
p r o b a b i l i t y o f f a l s e p o s i t i v e s and f a l s e n e g a t i v e s necessarily
exceeds t h a t f o r t h e i n d i v i d u a l peak d e t e c t i o n ( o r o u t l i e r detec-
tion) test. F o r example, i f a l a r g e gamma r a y spectrum c o n t a i n i n g
no a c t u a l r a d i o a c t i v i t y were scanned w i t h t h e e q u i v a l e n t o f , s a y ,
50 d e t e c t i o n d e c i s i o n s [a - 0.05], t h e r e a l m o s t c e r t a i n l y [>92%
chance] would be a t l e a s t one f a l s e p o s i t i v e peak. Similar
c o n s i d e r a t i o n s a p p l y t o f a l s e n e g a t i v e s , so f a l s e alarms and m i s s e d
r a d i o a c t i v i t y would be t h e consequence. ( I g n o r i n g t h i s i s s u e has
l e d t o some d i f f i c u l t i e s i n t h e e v a l u a t i o n o f l o w - l e v e l gamma r a y
s p e c t r a ; see R e f . 28 f o r f u r t h e r d i s c u s s i o n . ) The s o l u t i o n i s t o
f o l l o w t h e r u l e s f o r combining p r o b a b i l i t i e s ; namely, a d j u s t i n g t h e
s i g n i f i c a n c e l e v e l s so t h a t t h e o v e r a l l p r o b a b i l i t i e s o f c o r r e c t
non-detection [1-a'] and c o r r e c t d e t e c t i o n [l-fi'] remain 95%. The
p r o b a b i l i t y that a l l decisions are c o r r e c t i s simply the continued
n
p r o d u c t : (l-7r') = 0.95 - II(l-7r) = ( l - 7 r ) , where n r e p r e s e n t s a o r
P, and n, t h e e q u i v a l e n t number o f t e s t s p e r spectrum. Adjusted
v a l u e s f o r a and p a r e then g i v e n by (Eq. 5 ) ,
1 11
a (or p) = 1 - (0.95) / (5)
I f t h e t o t a l e r r o r l e v e l i s t o be h e l d a t 5% [ a ' , P' ] f o r a
m u l t i t e s t experiment i n w h i c h H i s a c t u a l l y t r u e 50 times and H ,
0 A
3 t i m e s , t h e n Eq. 5 g i v e s a d j u s t e d v a l u e s o f a = 0.00103, p = 0.017

w i t h c o r r e s p o n d i n g c r i t i c a l l e v e l s and d e t e c t i o n l i m i t s o f 3.1 o Q

and 5.2 a , respectively. Monitoring of mostly empty spectra thus

Q
provides j u s t i f i c a t i o n f o r unequal a,fi,and thus f o r Lp/I^ < 2.

3.3.4 multichannel i d e n t i f i c a t i o n . The linkage between
detection and i d e n t i f i c a t i o n was brought up e a r l i e r , where
" i d e n t i f i c a t i o n " was formulated i n terms of the s t a t i s t i c a l estima-
t i o n of the c h a r a c t e r i s t i c value (e.g., element concentration or
r a t i o , gamma ray energy) f o r the i d e n t i f y i n g v a r i a b l e . Linear
estimation was at l e a s t implied i n that discussion, so that
i n i t i a l l y normal data would lead to normal (though possibly
correlated) errors f o r the estimated r e s u l t s . For example, the
frequency d i s t r i b u t i o n of events (counts) along the energy axis
[ i d e n t i f y i n g variable] could be used to estimate the mean energy
("centroid") and i t s variance f o r a gamma ray peak, and the peak
magnitude or "area" could be simultaneously estimated with a simple
f i l t e r function to compensate for a l i n e a r baseline. The decision
space i s now two dimensional so contours of the b i v a r i a t e area
(detection) - energy ( i d e n t i f i c a t i o n
for significance testing
non-normal, or when they are subjected to non-linear operations as
i n c e r t a i n peak search and peak f i t t i n g algorithms, normality i s
not preserved, so caution i s i n order i n making detection decisions
and i n deriving confidence i n t e r v a l s .
" N o n - s t a t i s t i c a l " i d e n t i f i c a t i o n i s important i n many facets
of a n a l y t i c a l science, where signal l o c a t i o n or " i d e n t i t y b i n " pre-
determines species i d e n t i t y . Detection and i d e n t i f i c a t i o n are then
uncoupled, and any signal detected i n the c h a r a c t e r i s t i c b i n
simultaneously conveys detection and i d e n t i t y . C l a s s i c a l a n a l y t i -
c a l chemistry (e.g., gravimetry) r e l i e d heavily on t h i s model,
where unique chemical separations would guarantee i d e n t i t y . Modern
instrumental or chromatographic methods s i m i l a r l y succeed when the
resolving power (discrimination power) f a r exceeds the "density" of
pure components along the informing variable.
3.3.5 pattern discrimination limits [multivariable
identificationl. We considered the discrimination of chemical
components or classes e a r l i e r from a univariate perspective,
including the paired comparison f o r a single alternative or
contaminating component which would necessarily l i e to one side of
the n u l l class (known component against which the sample i s to be
compared). Before considering discrimination with multiple
chemical variables (compositional or spectral patterns), l e t us
broaden the univariate problem to two-sided discrimination, since
unlike analyte detection, c h a r a c t e r i s t i c or i d e n t i f y i n g variable
values generally may be larger or smaller than that of H . The H Q Q
discrimination l i m i t test would then be 2-sided -- i . e . , z - 1.96 c
instead of 1.645 for a - 0.05. I f a paired comparison of a test

sample (unknown) with the control sample (known, H ) f a l l s within Q
1.96 o , we then conclude that there i s no s t a t i s t i c a l l y sig-

0
n i f i c a n t difference. This i s not, however, proof that the patterns

are the same; i t i s only a test of consistency. I t i s necessary,
but not s u f f i c i e n t . To establish a r e a l match, or " i d e n t i f i c a -
t i o n , " we must demonstrate that the universe of alternative
patterns w i l l not match ( s t a t i s t i c a l l y ) . Design of a measurement
process f o r the successful i d e n t i f i c a t i o n of a p a r t i c u l a r chemical
species or compositional state thus requires consideration of both
a andfierrors, as depicted i n F i g . 12.

H' A
Mole Fraction
F i g . 11. Impurit
d e t e c t a b l e c o n c e n t r a t i o n o f substance-A [ H ] i n t h e " n u l l " sub-
A
s t a n c e [ H ] . The a b s c i s s a r e p r e s e n t s mole f r a c t i o n o r m i x i n g
0
ratio. I n d i v i d u a l i m p u r i t y d e t e c t i o n l i m i t s would o b t a i n f o r each

i m p u r i t y t y p e , e.g., A' [ H ' ] .A
TRUE MATCH
[PROB = 1-a = 0.90]
SI M
A -
B B < 9'9)
FALSE MATCH
[PROB = ft = 0.26]
A -B
A < ' "HlW
Si
F i g . 12. S i n g l e S p e c i e s M a t c h i n g ; U n i v a r i a b l e I d e n t i f i c a t i o n .
For a g i v e n l o c a t i o n on t h e a b s c i s s a [ i d e n t i f y i n g v a r i a b l e : i s o t o p e
r a t i o , X - r a y e n e r g y , . . . ] , unique i d e n t i f i c a t i o n r e q u i r e s t h a t none
o f t h e p o s s i b l e H 's o v e r l a p s ( p r o b a b i l i t y ft o r l e s s ) t h e t w o - s i d e d
A
H Q window [ I ] . That i s , a l l s e p a r a t i o n s must exceed t h e c o r -

responding d i s c r i m i n a t i o n l i m i t s . (From t h e d e s i g n p e r s p e c t i v e ,
s i n c e i d e n t i f y i n g v a r i a b l e s e p a r a t i o n s a r e g e n e r a l l y f i x e d by
N a t u r e , we must d e s i g n t h e CMP t o a c h i e v e c o r r e s p o n d i n g A ' s -- c f , D
F i g . 8A [ROC c u r v e ] . ) [ I l l u s t r a t i o n c o n s t r u c t e d u s i n g a= 0.10, A
= -4.0 mg/g, <7 1.0 mg/g, and a^ = 2.6 mg/g.]
fi

I t i s a s m a l l step t o t a k e from u n i v a r i a b l e i d e n t i f i c a t i o n to
m u l t i v a r i a b l e or p a t t e r n matching. I f we are c o n c e r n e d w i t h j u s t a
s i n g l e a l t e r n a t i v e p a t t e r n [ A ] , but s e v e r a l (n) measured v a r i a b l e s ,
t h e n the c o n s i s t e n c y t e s t r e q u i r e ; t h a t a l l n v a r i a b l e s match
s t a t i s t i c a l l y when the i d e n t i t y o f t e s t sample i s the same as t h a t
o f the c o n t r o l sample [B] . Combining p r o b a b i l i t i e s as b e f o r e ,
n
(1-a') = 0.95 = 11(1-0^) - ( l - a ) . P r o o f o f i d e n t i t y , as b e f o r e ,
includes c o n s i d e r a t i o n of s u f f i c i e n c y i . e . , we r e q u i r e i n
a d d i t i o n t h a t [A] not match ( s t a t i s t i c a l l y ) [B] s i m u l t a n e o u s l y f o r
a l l measured v a r i a b l e s . P r o b a b i l i t i e s are combined a little
d i f f e r e n t l y i n t h i s case; the o v e r a l l p r o b a b i l i t y o f an erroneous
match i s g i v e n by fi' = U(fi ). i The p r o d u c t i s a l s o t a k e n over a l l n
v a r i a b l e s , whose i n d i v i d u a l fi 's w i l l g e n e r a l l y d i f f e r .
L U n l e s s fi'
< 0.05, m a t c h i n g o f p a t t e r n s cannot e s t a b l i s h i d e n t i t y . A t the
same time, i t i s t h i s m u l t i p l i c a t i v e f e a t u r e , when i n d i v i d u a l fi's
are themselves small that gives multivariable or pattern
d i s c r i m i n a t i o n i t s enormou
To i l l u s t r a t e , l e
p a t t e r n s i n two pure source m a t e r i a l s , where the o r i g i n o f one
( c o n t r o l sample, B) i s known, as i s the c o m p o s i t i o n o f the p o s s i b l e
a l t e r n a t i v e A. G i v e n the c h a r a c t e r i s t i c s o f the measurement
p r o c e s s and the c o m p o s i t i o n s o f the two known s o u r c e s , we can t e l l
a p r i o r i whether the s o u r c e s are d i s c r i m i n a b l e as i n d i c a t e d above.
I f n o t , the c a p a b i l i t y o f an unknown t e s t sample t o match proves
nothing. Absence o f a match under t h e s e c o n d i t i o n s , however, would
d e s e r v e s c r u t i n y ; i t c o u l d i n d i c a t e e i t h e r f a u l t y measurements or
f a u l t y assumptions. I l l u s t r a t i v e d a t a are g i v e n i n T a b l e I I I .
Table I I I . Multivariable Identification
I n p u t d a t a f o r e s t i m a t i n g the d i s c r i m i n a b i l i t y ( i d e n t i f i a b i l i t y )
a c
o f p a r t i c l e e m i s s i o n s from s t e e l p l a n t s A and B ( >b> )
BvsS IV A vs 6
Al Si Ca Cr Mn Fe
Concentration (mg/g)
steel-B 10 12 45 3.2 22 160
steel-A 13 8 70 3.3 16 120
a 1. 1 1.0 5.8 0.32 1.9 14
window [ I ] 4. ,00 3.63 21.1 1.16 6.91 50.9
distance [A] 3.,0 -4.0 25. 0.10 -6.0 -40.
fi 0..74 0.40 0.32 0.98 0.63 0.71
1,.93 -2.83 3.05 0.22 -2.23 -2.02
(a) Based on d a t a from Ref. 72.

(b) V a l u e s o f I and fi are g i v e n f o r n=5.
(c) F i g . 13 d e p i c t s the windows [ I ] and v a r i a b l e s e p a r a t i o n s [A].

Concentrations f o r s i x elements c h a r a c t e r i z i n g two s t e e l

a e r o s o l samples (72.) a r e g i v e n i n t h e f i r s t two rows. S t e e l - B i s
t a k e n as t h e c o n t r o l , and s t e e l - A as t h e a l t e r n a t i v e s o u r c e . H i s 0
r e p r e s e n t e d by t h e v e c t o r o r p a t t e r n d i f f e r e n c e , (x - x ) ; H , by
B B A
(x - x ) . The l a s t f i v e rows o f t h e t a b l e i n d i c a t e , r e s p e c t i v e l y :
A B
the s t a n d a r d d e v i a t i o n s [a] f o r t h e elements i n q u e s t i o n , t h e

m a t c h i n g i n t e r v a l s [ I ] , t h e c o n c e n t r a t i o n d i f f e r e n c e s [A] under H , A
the p r o b a b i l i t y o f f a l s e matches [fi], and t h e r a t i o s o f c o n c e n t r a -

t i o n d i f f e r e n c e s [A] t o t h e p a i r e d measurement s t a n d a r d d e v i a t i o n s
[o ] .
0 1-0 and A / a b o t h s e r v e as measures o f i n d i v i d u a l element
Q
d i s c r i m i n a t i n g power. The q u a n t i t y I i s computed by r e q u i r i n g 1-a'

t o be 0.95; f o r n=5, t h i s means a - 0.0102 o r z ( 2 - s i d e d ) = 2.57. c
(For 6-member p a t t e r n s , z i n c r e a s e s t o 2.63.) Then I = z c r ,

c c Q
where a = ajl.
Q P a t t e r n d i f f e r e n c e s [A] , i n d i c a t e d by t h e open
c i r c l e s , a r e shown i n comparison w i t h m a t c h i n g i n t e r v a l s i n F i g .
13.
For t h i s example
has been approached i
the p r o d u c t o f t h e i n d i v i d u a l 's, r e f l e c t i n g the series of
i n d i v i d u a l element matching d e c i s i o n s . ( F o r n = 5, o m i t t i n g Ca,
t h i s p r o d u c t e q u a l s 0.13.) Second, t h e v e c t o r d i f f e r e n c e r e p r e -
2
s e n t e d by H i s examined t h r o u g h t h e use o f t h e n o n - c e n t r a l
A x
2 n
s t a t i s t i c , where S ( A / a ) i s t h e n o n - c e n t r a l i t y parameter (7J3) . *
Q
t h i s second case t h e t e s t o f t h e v e c t o r match ( i . e . , H t e s t ) i s Q
c a r r i e d o u t by comparing t h e sum o f squares o f t h e n observed

2
normalized differences with the c r i t i c a l l e v e l f o r the c e n t r a l x
f o r n - degrees o f freedom. The rms v a l u e from t h e sum o f squares
-- ( A / a ) Q r i n s-- r e p r e s e n t s t h e m u l t i v a r i a b l e g e n e r a l i z a t i o n o f t h e
u n i v a r i a t e normalized differences. I t i s a convenient s i n g l e
parameter measure ( i n d e x ) f o r t h e v e c t o r d i s c r i m i n a t i o n power
(1-/?'), as f$' i s u n i q u e l y d e t e r m i n e d by t h i s q u a n t i t y , g i v e n a' and
the number o f degrees o f freedom.
T a b l e IV g i v e s r e s u l t s f o r t h e two t y p e s o f t e s t and s e v e r a l
c h o i c e s o f element p a t t e r n s . Important d u a l p a t t e r n i d e n t i f i c a t i o n
conclusions follow: (a) D i s c r i m i n a t i o n power ( i d e n t i f i a b i l i t y )
2
d i f f e r s a c c o r d i n g t o t h e type o f t e s t , x being significantly
b e t t e r and becoming more so w i t h i n c r e a s e d d i m e n s i o n a l i t y . (b)
O p t i m a l f e a t u r e s e l e c t i o n (e.g., f o r n=5) g i v e s o p t i m a l d i s c r i -
m i n a t i n g power f o r t h e number o f v a r i a b l e s s e l e c t e d . ( c ) There
e x i s t s an o p t i m a l number o f dimensions ( v a r i a b l e s ) . The most
p o w e r f u l v a r i a b l e (here, Ca) i s used f o r n1; a second d i s c r i m i -
n a t i n g v a r i a b l e y i e l d s i n c r e a s e d power w i t h n=2; b u t e v e n t u a l l y
a d d i t i o n o f poorly d i s c r i m i n a t i n g v a r i a b l e s " d i l u t e s " the d i s c r i -
m i n a t i o n power -- e.g., n=6 compared t o t h e b e s t s e t o f 5. (d)
I n c r e a s e d d i m e n s i o n a l i t y g i v e s enormous l e v e r a g e t o modest improve-
ments i n p r e c i s i o n , through t h e p r o d u c t ILp . i (See bottom l i n e ,
T a b l e IV.) These f o u r c o n c l u s i o n s d i r e c t l y i n d i c a t e t h e way toward
improved d i s c r i m i n a t i o n power, t h e l a s t b e i n g t h e most i n f l u e n t i a l .
2
(X 2
i n t h e t a b l e denotes t h e n o n - c e n t r a l x )
3.3.6 generalization. The f o r e g o i n g considerations of
h y p o t h e s i s t e s t i n g and p a t t e r n i d e n t i f i c a t i o n l i m i t s were neces-
s a r i l y s i m p l i f i e d , an extended d i s c u s s i o n b e i n g beyond t h e scope o f

Library
115516th st, N.w.Currie, L.;
In Detection in Analytical Chemistry;
ACS Symposium Series; Washington, D.C.
American Chemical 20036
Society: Washington, DC, 1987.
+6
O)
E +4
+2
5 0 1
Element
Al Si Cr Mn Fe Ca
(1/10) (1/5)
S -2 I
I
I 1
F i g . 13. M u l t i v a r i a b l e I d e n t i f i c a t i o n . H windows [ I ] and H

Q A
concentration differences [open c i r c l e s ] f o r the m u l t i - v a r i a b l e

(element) p a t t e r n s c h a r a c t e r i s t i c o f p a r t i c l e e m i s s i o n s from two
s t e e l p l a n t s . [See T a b l e I I I . ]

T a b l e IV. P a t t e r n D i s c r i m i n a t i o n Power [1 - P']
(Steel-A p a r t i c l e s vs Steel-B p a r t i c l e s [ c o n t r o l ] ; a' - 0.05)
n-1 n=5 n=6

(Ca) (-Ca) <-Cr)
Sequential m a t c h i n g ^ 0.86 0.87* 0.958 0.949

2
X - test< ) b
0.88 0.967 0.997 0.996
(a) c r i t e r i o n : x, < I, , a l l i power: 1 - II 0
(b) c r i t e r i o n : x ^ Xc power: 1 - P(x' ) 2
*8% i n c r . p r e c i s i o n ( o 's) i i n c r e a s e s t h e power t o 0.95 [ t a r g e t ]
t h i s chapter. The p r i n c i p a l g e n e r a l i z a t i o n s t h a t s h o u l d be
c o n s i d e r e d , however, a r e t h e f o l l o w i n g :
(1) F o r t h e f i r s t ("matching") s t r a t e g y , t h e requirement o f
homogeneous v a r i a n c e may be r e l a x e d w i t h t h e use o f i n d i v i d u a l o's:
i . e . , o j2 f o r t h e e v a l u a t i o n o f a and I , and 7 ( a
B
z
+ a ) , to
A
z
B
r e c a l c u l a t e t h e fi (See F i g u r e 1 2 ) .

2
(2) F o r v a r i a n c e s e s t i m a t e d as s ' s , t and F would r e p l a c e z
2
and x i r e s p e c t i v e l y , f o r h y p o t h e s i s t e s t i n g . To e s t i m a t e t h e
power o f t h e t e s t s , t h e c o r r e s p o n d i n g n o n - c e n t r a l d i s t r i b u t i o n s
w o u l d be employed. The n o n - c e n t r a l i t y parameter f o r t h e F
2
d i s t r i b u t i o n i s t h e same as f o r x T h i s means t h a t even i n t h e
b e s t o f c i r c u m s t a n c e s ( o r t h o g o n a l v a r i a b l e s ) t h i s approach t o t h e
i d e n t i f i c a t i o n l i m i t o r power r e q u i r e s homogeneity o f v a r i a n c e and
knowledge o f o. (See r e f e r e n c e (74) f o r a d i s c u s s i o n o f these
i s s u e s , as w e l l as an i n - d e p t h t r e a t m e n t o f m u l t i v a r i a t e h y p o t h e s i s
t e s t i n g and c l a s s i f i c a t i o n . )
(3) I f t h e H u n i v e r s e c o n t a i n s more t h a n one member, i t s
A
membership and c o m p o s i t i o n must be known f o r i d e n t i f i c a t i o n t o be

meaningful. Such knowledge, o f c o u r s e , i s i n t h e domain o f
disciplinary ("scientific") expertise. Proof o f H Q[identifica-
t i o n ] comes o n l y when d i s c r i m i n a t i n g power i s adequate w i t h r e s p e c t
to a l l H ' s . A F o r a g i v e n c o n t r o l p a t t e r n B, o n l y t h a t r e g i o n o f
v a r i a b l e space w i t h i n t h e d i s c r i m i n a t i n g volume need be e x p l o r e d ,
however. F o r s e q u e n t i a l m a t c h i n g , t h i s means o n l y A - p a t t e r n s f o r
w h i c h t h e d i s t r i b u t i o n o f t h e d i f f e r e n c e spectrum A - B s i g -
n i f i c a n t l y o v e r l a p s the I-hypercube; f o r t h e a l t e r n a t i v e approach,
2
the d i s c r i m i n a t i n g volume d e r i v e s from t h e c r i t i c a l v a l u e f o r x
M u l t i c l a s s d i s c r i m i n a t i o n may be performed, f o r example, through a
s e r i e s o f b i n a r y t e s t s (12, T7, 7 4 ) .
(4) I m p u r i t y d e t e c t i o n f o r t h e m u l t i v a r i a b l e case may be
t r e a t e d as a d i r e c t e x t e n s i o n o f t h e s i n g l e v a r i a b l e case. F o r two
p a t t e r n s , t h e i m p u r i t y d e t e c t i o n l i m i t (component A c o n t a m i n a t i n g
c o n t r o l component B) c a n be c a l c u l a t e d from ( A / a ) 0 r m s corresponding
2
to P' - 0.05, where x (<*' - 0.05) i s u s e d t o t e s t t h e n u l l
hypothesis [H : x Q B - & ].B F o r mixed i m p u r i t i e s , a "worst case"

l i m i t may be d e r i v e d from the " p s e u d o p a t t e r n " o f c l o s e s t approach.

That i s , the two pattern discrimination l i m i t i s recalculated
s u b s t i t u t i n g A' f o r A, where p s e u d o p a t t e r n A' i s the l i n e a r
c o m b i n a t i o n o f a l t e r n a t i v e v e c t o r s w h i c h l i e s c l o s e s t t o B. At
t r a c e l e v e l s , o b s e r v e d p a t t e r n s become i n c r e a s i n g l y f u z z y , because
o f measurement i m p r e c i s i o n o r b a s e l i n e n o i s e . C l e a r l y , under such
circumstances "detection" and " i d e n t i f i c a t i o n " become entwined.
(See r e f e r e n c e (75) .)
(5) Covariance among variables (e.g., elemental
c o n c e n t r a t i o n s ) w i t h i n i n d i v i d u a l c l a s s e s i s by f a r the greatest
complication. I t may be t r e a t e d i n one o f f o u r ways: (a) S e l e c t
j u s t one v a r i a b l e o r f u n c t i o n o f v a r i a b l e s ( e . g . , f i r s t p r i n c i p a l
component); c o v a r i a n c e i s t h e n u n d e f i n e d . (b) S e l e c t o n l y the most
powerful, uncorrelated (or nearly independent) variables,
d i s c a r d i n g o t h e r s showing s i g n i f i c a n t c o r r e l a t i o n (12, Chapt 20).
(c) T r a n s f o r m the o r i g i n a l v a r i a b l e s i n t o a r e d u c e d orthogonal
s e t , as i n P r i n c i p a l Componen
(d) I n the absence o f
v a r i a t e normal a s s u m p t i o n and e s t i m a t i n g the w i t h i n c l a s s c o v a r i -
ance m a t r i c e s , the f o u r t h a l t e r n a t i v e i s d a u n t i n g : t a k i n g i n t o
a c c o u n t the f u l l c o v a r i a n c e s t r u c t u r e t h r o u g h c r i t i c a l c o n t o u r s
[a,fi] o f the h y p e r e l l i p s o i d s c o r r e s p o n d i n g t o H and H . Q A Consider-
i n g j u s t two v a r i a b l e s , the t r e a t m e n t would be analogous t o the
c o n f i d e n c e e l l i p s e f o r the e s t i m a t e d s l o p e and i n t e r c e p t o f a
fitted calibration line. (Hypothesis t e s t i n g of c a l i b r a t i o n curve
p a r a m e t e r s i s f a r more amenable t o t h i s m u l t i v a r i a t e " p a r a m e t r i c "
approach, however, s i n c e the c o r r e l a t i o n m a t r i x i s known from the
d e s i g n o f the e x p e r i m e n t (77) .) For two v a r i a b l e s , the m a t c h i n g
i n t e r v a l s I and the r e s p e c t i v e p r o b a b i l i t i e s (1-a') would n o t be
greatly a f f e c t e d by the lack of rigorous knowledge o f the
2
c o v a r i a n c e m a t r i x , s i n c e (1-a') ( 1 - a ) ~ 1. The f a l s e match
p r o b a b i l i t y fi' c o u l d be s i g n i f i c a n t l y i n e r r o r , however, because
must now be r e p l a c e d by P *(0 \1),
1 2 where ( |1) i s the
2
c o n d i t i o n a l p r o b a b i l i t y o f a f a l s e match f o r v a r i a b l e - 2 g i v e n a
f a l s e match f o r v a r i a b l e - 1 . I f the v a r i a b l e s are p e r f e c t l y
c o r r e l a t e d , ( |1) - 1, and the second v a r i a b l e l e n d s no i n c r e -
2
m e n t a l d i s c r i m i n a t i n g power. H i g h e r d i m e n s i o n s l e a d t o i n c r e a s i n g
c o m p l e x i t y , and e s t i m a t e s o f h i g h e r o r d e r c o r r e l a t i o n s become
i n c r e a s i n g l y i m p r e c i s e as one runs out o f degrees o f freedom.
4. CONCLUSIONS AND OUTLOOK
The a b i l i t y to detect s p e c i f i e d (absolute) l e v e l s of chemical

s p e c i e s i n e n v i r o n m e n t a l , b i o l o g i c a l , and p h y s i c a l ( m a t e r i a l ) sys-
tems i s c r u c i a l f o r the w e l l - b e i n g and advancement o f our s o c i e t y .
Because o f the p r a c t i c a l importance o f r e l i a b l e d e t e c t i o n i n the
s o c i e t a l s e t t i n g , on the one hand, and i t s t e c h n i c a l c o m p l e x i t y , on
the o t h e r , we f a c e a "Two C u l t u r e s " type s i t u a t i o n . We s c i e n t i s t s
l a c k the e x p e r t i s e t o f u l l y comprehend o r e f f e c t i v e l y i n f l u e n c e the
s o c i o p o l i t i c a l i s s u e s ; e x p e r t s i n t h a t domain, s i m i l a r l y cannot be
expected to fully comprehend the technical issues involved.
E f f e c t i v e communication and m u t u a l e d u c a t i o n - - one o f the aims o f
t h i s volume and t h i s o v e r v i e w c h a p t e r -- i s t h e r e f o r e e s s e n t i a l .
W i t h t h i s o b j e c t i v e i n mind, l e t us r e - c o n s i d e r b r i e f l y some o f the
o b s e r v a t i o n s and s u g g e s t i o n s o f t h i s t r i p a r t i t e o v e r v i e w .

4.1 Sociopolitical Perspectives

o Adequate d e t e c t i o n c a p a b i l i t i e s are i m p o r t a n t t o s o c i e t y , b o t h
for natural or anthropogenic hazards and for requisite
b e n e f i c i a l l e v e l s of chemical species e.g., n u t r i e n t s .
"Adequacy" means " c e r t a i n t y " t o the layman; i f a substance i s
p r e s e n t (above the s p e c i f i e d l e v e l o f concern) i t w i l l s u r e l y be
d e t e c t e d -- the alarm w i l l go o f f ; i f n o t , t h e r e w i l l be no
( f a l s e ) alarm.
o D e s p i t e the i n t r i n s i c u n c e r t a i n t y ( f a l s e p o s i t i v e s and f a l s e
n e g a t i v e s ) a s s o c i a t e d w i t h d e t e c t i o n , and i n f a c t , w i t h a l l o f
measurement, the g e n e r a l p u b l i c i s not s c h o o l e d t o a c c e p t such a
limitation. Ignorance and s u s p i c i o n w i t h r e s p e c t t o t h i s i s s u e
i s r e f l e c t e d a l s o when i t comes t o our a b i l i t y t o reduce
c o n c e n t r a t i o n s o f "bad" s p e c i e s t o z e r o , or f o r t h a t m a t t e r t o
d e t e c t a l l concentrations exceeding zero (78) Scientific
n a i v e t e ' r e g a r d i n g newl
c a p a b i l i t i e s improv
h a v i n g p o t e n t i a l l y g r e a t p o l i t i c a l impact.
o S o c i o p o l i t i c a l "debates," i n b o t h the l e g i s l a t i v e and j u d i c i a l
arenas, have v e r y d i f f e r e n t ground r u l e s t h a n s c i e n t i f i c debates
(3.). Advocacy, c o n f l i c t i n g s o c i e t a l concerns and p e r c e p t i o n s ,
and even "hidden agenda" d r i v e such debates. They cannot, and
p r o b a b l y s h o u l d n o t , be conducted l i k e a s c i e n t i f i c forum. W i t h
p a t i e n c e and h o n e s t i n p u t from the s c i e n t i f i c community i n i t s
a r e a o f e x p e r t i s e , g e n e r a l l y the c o l l e c t i v e common good i s
s e r v e d (79.80). W i t h r e f e r e n c e t o r i s k management f o r "dread
r i s k s " a f f e c t i n g l a r g e numbers o f p e o p l e , f o r example, Lave
o b s e r v e d t h a t c o l l e c t i v e d e c i s i o n s are mandatory, b u t because o f
the diversity of safety goals, collective decisions are
d i f f i c u l t (80).
o R i s k p e r c e p t i o n s and c o l l e c t i v e ( o r d e l e g a t e d ) d e c i s i o n s l i e
b e h i n d many o f our r e g u l a t o r y l i m i t s o r h a z a r d "alarm l e v e l s , "
[1^] w h i c h , i n t u r n , d r i v e our measurement D e t e c t i o n L i m i t s .
Though c e r t a i n approaches t o d e c i s i o n a n a l y s i s , e s p e c i a l l y those
i n c o r p o r a t i n g B a y e s i a n s t r a t e g i e s , might seem a p p r o p r i a t e f o r
s i m u l t a n e o u s l y embracing s o c i e t a l r i s k and measurement e r r o r
r i s k ( f a l s e p o s i t i v e s and n e g a t i v e s ) , i t would seem a d v i s a b l e i n
p r a c t i c e t o decouple the two. Let s o c i e t y (or medicine, or
a f f e c t e d i n d u s t r y , e t c . ) e n t e r the p o l i t i c a l debate t o e s t a b l i s h
t h e i r r e q u i s i t e L r ' s. Then, Measurement S c i e n c e , u s i n g the
a p p r o p r i a t e s c i e n t i f i c c r i t e r i a and s t a n d a r d s , s h o u l d attempt t o
meet these 1^ 's w i t h s c i e n t i f i c a l l y d e f e n s i b l e D e t e c t i o n L i m i t s .
The l a t e P h i l i p H a n d l e r put i t w e l l , by s t a t i n g t h a t " S c i e n t i s t s
b e s t s e r v e p u b l i c p o l i c y by l i v i n g w i t h i n the e t h i c s o f s c i e n c e ,
not those o f p o l i t i c s " ( 8 1 ) .
o S o c i e t a l and s c i e n t i f i c p e r c e p t i o n s o f r i s k sometimes d i v e r g e .
S l o v i c ' s i n v e s t i g a t i o n o f o r d e r e d r i s k p r e f e r e n c e s o f laymen vs
e x p e r t s i s an i n t e r e s t i n g case i n p o i n t ( 7 9 ) . N u c l e a r power,
f o r example, was r a t e d f i r s t among r e p r e s e n t a t i v e s o f the l a y
p u b l i c (League o f Women V o t e r s ; c o l l e g e s t u d e n t s ) , y e t i t was
2 0 t h i n the eyes o f e x p e r t s . S u r g e r y was 5 t h a c c o r d i n g t o the
r e l e v a n t e x p e r t s , but i t was o n l y 10th i n the p u b l i c view. The
n a t u r e o f our s o c i e t y n a t u r a l l y a c c o r d s p r i m a r y w e i g h t t o t h a t

DETECTION IN ANALYTICAL CHEMISTRY
society's public perceptions, when i t comes t o political

decisions. T h i s d e s e r v e s our r e s p e c t f o r many r e a s o n s , i n c l u d -
i n g the f a c t t h a t s o c i e t y ' s judgment i s not c o n s t r a i n e d by a
p o s s i b l y too narrow v i e w o r f a u l t y a l g o r i t h m . In fact, i t s
" b a s i c c o n c e p t u a l i z a t i o n o f r i s k i s much r i c h e r t h a n t h a t o f the
e x p e r t s and r e f l e c t s l e g i t i m a t e concerns t h a t are typically
o m i t t e d from e x p e r t r i s k assessments" (79, p. 285).
Adequacy o f d e t e c t i o n l i m i t s i s something t h a t s o c i e t y has a
r i g h t t o demand, and s u p p o r t i f the c o s t s h o u l d be high.
Inadequate d e t e c t i o n c a p a c i t y f o r s p e c i f i e d l e v e l s o f f i r e s ,
e a r t h q u a k e s , t o x i c organisms, e t c . must be a d d r e s s e d through
r e f i n e d s a m p l i n g and measurement p r o c e d u r e s . Inadequate p e r f o r -
mance o f a Measurement P r o c e s s not o n l y f a i l s t o provide
s u f f i c i e n t w a r n i n g , b u t i t may a l s o produce q u i t e m i s l e a d i n g
conclusions. E l e v a t e d l e v e l s o f N i i n human serum due to
o c c u p a t i o n a l exposur
undetectable u n t i l a
d e v e l o p e d under the a u s p i c e
and A p p l i e d C h e m i s t r y [IUPAC]. P r i o r methods, q u i t e i n c o r r e c t l y
i m p l i e d t h a t normal l e v e l s o f N i i n b l o o d serum were some t e n
t i m e s h i g h e r t h a n t h a t o c c u p a t i o n a l exposure l e v e l ( 8 2 ) .
The costs o f erroneous d e t e c t i o n d e c i s i o n s can be quite
significant. D i s a s t r o u s r e s u l t s may follow i f irreversible
a c t i o n s are taken. Even the s e e m i n g l y h a r m l e s s f a l s e p o s i t i v e
w h i c h can l a t e r be shown t o be s p u r i o u s can damage r e p u t a t i o n s
and/or l e a d t o e x p e n s i v e c o u r t s u i t s . I t i s important therefore
t h a t s c i e n t i f i c d e t e c t i o n d e c i s i o n s and d e t e c t i o n l i m i t s be
approached i n a q u a n t i t a t i v e manner, w i t h due a t t e n t i o n t o the
p r o b a b i l i t y of e r r o r s of both kinds.
.2 Technical Issues
M e a n i n g f u l d e t e c t i o n d e c i s i o n s and d e t e c t i o n l i m i t s can f o l l o w
o n l y from r i g o r o u s a t t e n t i o n t o the Measurement P r o c e s s and an
H y p o t h e s i s T e s t i n g framework f o r d e f i n i n g d e t e c t i o n c a p a b i l i t y .
T h i s i s e s p e c i a l l y a p p r o p r i a t e , as h y p o t h e s i s t e s t i n g i s the
expression of the S c i e n t i f i c Method. Decision criteria,
d e t e c t i o n l i m i t s , and acceptable f a l s e p o s i t i v e and false
n e g a t i v e r i s k s must be q u a n t i f i e d , and CMP's d e s i g n e d t o meet
t h e i r s p e c i f i c a t i o n s . The s c i e n t i f i c e x p e r t i s e r e q u i r e d goes
deep. T h i s was o b s e r v e d , f o r example, i n the i n v e s t i g a t i o n o f
d e t e c t i o n l i m i t s f o r a v a r i e t y o f a n a l y t i c a l methods f o r the
I n t e r n a t i o n a l Atomic Energy Agency. As i l l u s t r a t e d i n Ref. 35.
d e t a i l e d , method s p e c i f i c e x p e r t i s e was e s s e n t i a l i n o r d e r t o
expose c e r t a i n s u b t l e , b u t e x t r e m e l y i m p o r t a n t f a c t o r s a f f e c t i n g
c a l i b r a t i o n and the b l a n k fNote 71.
A l l i s not w e l l i n the t e c h n i c a l camp. C o n f u s i o n among
s c i e n t i s t s between the d e s i g n o f the MP t o meet r e q u i s i t e l e v e l s
o f performance [ 1 ^ ] , and an e x p e r i m e n t a l outcome o r d e t e c t i o n
d e c i s i o n b a s e d on a s p e c i f i e d c r i t e r i o n [ 1 ^ ] , i s a t the h e a r t o f
much our i n t e r n a l d i s a r r a y . That i s , two d i f f e r e n t ( a l b e i t
r e l a t e d ) i s s u e s are under d i s c u s s i o n , o f t e n unknowingly and w i t h
c o n f l i c t i n g terminology. Ad hoc r u l e s o f thumb, o r s i m p l i s t i c
consensus ("voted") formulae are p r o f f e r e d o f t e n i n the

CURRIE Overview of Historical, Societal, and Technical Issues
i n t e r e s t of producing a simple ranking of CMP's according to

something labeled as an LOD. This serves no one. In par-
ticular, i t f a i l s to provide the public with meaningful
detection c a p a b i l i t i e s comprising r e l i a b l e and adequate f a l s e
p o s i t i v e and f a l s e negative error p r o b a b i l i t i e s . Perhaps the
most common extreme i s the case where the fi-error is
unrecognized, such that i t s de facto value i s 50% fNote 81.
The drive toward f a c i l e expressions f o r l i m i t s of detection i s
p a r t l y a matter of attitude and education. Solid training i n
s t a t i s t i c s and d r i l l i n g with respect to the fundamental concepts
of experimental design and hypothesis testing i n science i s
missing from the undergraduate education of many chemists i n the
U.S. Western Europe fares better; and now that t r a i n i n g i n
Chemometries i s beginning to appear i n the American c u r r i c u l a
(2) r e a l hope exists f o r common understanding of these
matters by the "ordinary
t i o n of the present
an instrument manufacturer of i t s users i n the nuclear industry.
Regarding t o p i c a l material covered at workshops, comments came
back that users would prefer omission of the theory with more
time spent on use of the formulas. A personal view i s that
education related to basic concepts should always have the
priority; understanding (and questioning of) formulas is
important, but calculators or computers are quite p r o f i c i e n t at
using them.
The l i n k between "ordinary" measurements and detection l i m i t s

needs reinforcing. That i s , both depend for t h e i r v a l i d i t y on
a l l sources of systematic and random error associated with the
entire CMP. Thus, for example, detection decisions [tests of
significance] and confidence intervals depend on the same
assumptions and error components for their v a l i d i t y . If
Student's t i s appropriate to use with the one, i t i s equally
appropriate for the other.
Conventions for reporting data, and "black box" algorithms can
induce subtle bias into many types of modern chemical/
instrumental data, but the problem i s exacerbated with the
growth of automatic laboratory systems and low l e v e l measure-
ments and data bases. The black box may contain mistakes, and
a l l too often i t s mechanism i s unavailable to the user, and on
occasion that mechanism ( i . e . , algorithm) changes f o r low l e v e l
observations. Information loss or d i s t o r t i o n , whether i t occurs
within the black box or by the pen of the user, i s especially
severe for low l e v e l data. I t s impact on long term storage and
data base generation i s an issue of some importance (34).
Quality control at low levels (blank, detection l i m i t ) must be
addressed both with Standard Reference Materials and Standard
Test Data, i f we are to c e r t i f y the accuracy of our detection
decisions and detection l i m i t s . Since the blank has such a
profound influence on the v a l i d i t y of detection decisions, i t
deserves special attention. The CMP must be designed to incor-
porate an adequate number of " r e a l " blanks, and i t should take
advantage of the normalizing tendency of averages from paired
comparisons.

o The i n t r o d u c t i o n o f D i s c r i m i n a t i o n L i m i t s , such t h a t s m a l l
non-zero c o n c e n t r a t i o n s w i l l r a r e l y produce f a l s e p o s i t i v e s ,
s h o u l d do much t o a l l e v i a t e the p u b l i c a l a r m t h a t sometimes
f o l l o w s such " d e t e c t i o n . " A t the same time i t c o u l d a v e r t the
common i m p l i c i t overcompensation a s s o c i a t e d w i t h i g n o r i n g o f the
e r r o r o f the second k i n d [ f a l s e n e g a t i v e ] . A l s o , those who
d e c r y c u r r e n t usage o f d e t e c t i o n l i m i t s because they are too
i m p r e c i s e , o r e q u i v a l e n t t o the f l i p p i n g o f a c o i n , might r e g a r d
D i s c r i m i n a t i o n L i m i t s as u s e f u l , more p r e c i s e measures o f
detection capability, s t i l l i n k e e p i n g w i t h the hypothesis
t e s t i n g concept.
o D i s c r i m i n a t i o n L i m i t s and m u l t i p l e d e t e c t i o n d e c i s i o n s l e a d
naturally t o u n i v a r i a t e and m u l t i v a r i a t e formulations for
i d e n t i f i c a t i o n , an outgrowth o f the fundamental concept o f
hypothesis t e s t i n g . Methods f o r t r e a t i n g t h i s l i n k have been
d e v e l o p e d , so i t become
together the two primar
A n a l y s i s : D e t e c t i o n and I d e n t i f i c a t i o n .
o Identification differs i n one, very c r i t i c a l respect from
detection: a c o n s i s t e n c y t e s t o f the n u l l h y p o t h e s i s i s
n e c e s s a r y b u t not s u f f i c i e n t f o r i d e n t i f i c a t i o n . D i s c r i m i n a t i o n
l i m i t s must be adequate f o r a l l a l t e r n a t i v e hypotheses ( o t h e r
substances). At t h i s p o i n t s c i e n t i f i c i n t u i t i o n or e x p e r t i s e
p l a y s a c r u c i a l r o l e , f o r we must somehow d i s c o v e r the u n i v e r s e
o f a l l p o s s i b l e a l t e r n a t i v e s t o the substance we w i s h t o
i d e n t i f y , i n the c o n t e x t o f the g i v e n measurement p r o c e s s .
4.3 Pre- and P o s t - H i s t o r y : The C h a l l e n g e . The concept o f the

D e t e c t i o n L i m i t , a t l e a s t i n A n a l y t i c a l Chemistry, was slow t o
e v o l v e i n the e a r l y decades o f t h i s c e n t u r y from a l o o s e , q u a l i t a -
t i v e idea, to a p o t e n t i a l l y semi-rigorous numerical a t t r i b u t e f o r a
f u l l y - d e f i n e d CMP. D u r i n g the p a s t twenty y e a r s o r so, i m p o r t a n t
s t r i d e s have been made i n e d u c a t i o n and i n the development o f a
c o n s i s t e n t and p r a c t i c a l l y u s e f u l f o r m u l a t i o n o f the D e t e c t i o n
L i m i t , e s p e c i a l l y i n Europe. U n f o r t u n a t e l y , d i v e r s i t y i n under-
s t a n d i n g , f o r m u l a t i o n , and nomenclature among s c i e n t i s t s c o n t i n u e s .
T h i s has been e x a c e r b a t e d by the demand f o r r e g u l a t i o n s and
s i m p l i f i e d r u l e s and f o r m u l a s , o f t e n on r e l a t i v e l y s h o r t time
scales. " D e f i n i t i o n s " d e r i v i n g from p o l e m i c s o r from d e m o c r a t i c ,
consensus t a c t i c s are u n l i k e l y t o meet l o n g term s t a n d a r d s f o r
scientific rigor (conceptual rigor, not necessarily
u n c e r t a i n t y - f r e e , numerical rigidity).
A l t h o u g h a sound approach t o d e t e c t i o n has been a v a i l a b l e f o r
a t l e a s t two decades, and d e s p i t e i t s c u r r e n t s u c c e s s f u l a p p l i c a -
t i o n t o many p r a c t i c a l and s c i e n t i f i c problems, the c u r r e n t d i s -
a r r a y among s c i e n t i s t s i n the U.S. [ c f F i g . 4] can o n l y f u r t h e r
m y s t i f y the p u b l i c i n an a r e a t h a t seems a l r e a d y i n h e r e n t l y
mystical. The promise comes from t r e n d s i n c h e m i c a l e d u c a t i o n and
from work i n p r o g r e s s i n r e p u t a b l e i n t e r n a t i o n a l c h e m i c a l o r g a n i z a -
tions. S t a t i s t i c s and the p r o p e r c o n c e p t s o f measurement uncer-
t a i n t y , e x p e r i m e n t a l d e s i g n , and h y p o t h e s i s t e s t i n g are g a i n i n g a
f o o t h o l d i n the undergraduate c h e m i c a l c u r r i c u l u m , e s p e c i a l l y under
the s t i m u l i o f modern i n s t r u m e n t a l and c o m p u t a t i o n a l f a c i l i t i e s and
Chemometrics ( 8 4 ) . A l s o , a t the p r e s e n t time a t l e a s t two

commissions of IUPAC, partly i n collaboration with the

i n t e r n a t i o n a l chemometrics community, a r e d r a f t i n g g u i d e l i n e s and
nomenclature documents t r e a t i n g a b r o a d range o f c h e m i c a l measure-
ment i s s u e s , i n c l u d i n g those r e l a t e d t o u n c e r t a i n t y , experiment
d e s i g n , r e p o r t i n g o f d a t a , and d e t e c t i o n .
Cooperation between t h e two cultures should become
increasingly f r u i t f u l , as common c o n c e r n i n meeting s o c i e t y ' s
l e g i m i t a t e needs f o r p r a c t i c a l d e t e c t i o n c a p a b i l i t i e s b i n d us
t o g e t h e r , and as we each i n v e s t o u r e f f o r t s i n o u r r e s p e c t i v e areas
of e x p e r t i s e . M u t u a l e d u c a t i o n and i n t e r - c u l t u r a l communication
can o n l y a c c e l e r a t e t h i s p r o c e s s .
Acknowledgment
S p e c i a l thanks go t o t h e f o l l o w i n g c o l l e a g u e s , f o r t h e i r
i m p o r t a n t s u g g e s t i o n s and c a r e i n r e a d i n g a d r a f t o f t h i s c h a p t e r :
K. R. E b e r h a r d t , M. S
A. P e l l a , C. H. Spiegelman
Literature Cited
1. Kutschera, W. "Rare Particles"; Nucl. Instrum. Meth. B5,
1984, 233, 420.
2. Cooper, R. M. "Stretching Delaney T i l l It Breaks";
Regulation, Nov/Dec 1985, 11.
3. Moss, T. "Scientific Measurements and Data in Public Policy
Making"; Chapt. 3 in this volume.
4. Science, Risk Assessment Issue, 1987, 236, 267-300.
5. Currie, L.; Klouda, G.; Voorhees, K. "Atmospheric Carbon";
Nucl. Instrum. Meth., B5, 1984, 233, 371.
6. Rogers, L. B. "Interlaboratory Aspects of Detection Limits
Used for Regulatory/Control Purposes"; Chapt. 5 in this
volume.
7. Rensberger, B. "A Life is Worth $2 Million, Regulatory
Analysis Shows"; Science Notebook, Wash. Post, Mar. 2, 1987.
[Science news summarizing highlights of a study to be pub-
lished in Environmental Science and Technology.]
8. Liteanu, C.; Rica, I. Statistical Theory and Methodology of
Trace Analysis. New York: John Wiley & Sons; 1980.
9. Egan, J . P. Signal Detection Theory and ROC Analysis,
Academic Press, New York, 1975.
10. Frank, I. E.; Pungor, E.; Veress, G. E. "Statistical Decision
Theory Applied to Analytical Chemistry"; Anal. Chim. Acta 133
(1981) 433.
11. Howard, R. A. "Decision Analysis: Perspectives on Inference,
Decision, and Experimentation"; Proc. IEEE, 1970, 58, 823.
12. Massart, D. L.; Dijkstra, A . ; Kaufman, L. Evaluation and
Optimization of Laboratory Methods and Analytical Procedures,
New York, Elsevier, 1978.
13. Currie, L. A. The Discovery of Errors in the Detection of
Trace Components in Gamma Spectral Analysis, in Modern Trends
in Activation Analysis, Vol. I I . J . R. DeVoe; P. D. LaFleur,
Eds.; Nat. Bur. Stand. (U.S.) Spec. Publ. 312; p. 1215, 1968.

14. Kirchmer, C. J . "The Estimation of Limit of Detection for

Environmental Analytical Procedures" - Chapt. 4 in this
volume.
15. Boumans, P. W. J . M. A Tutorial Review of Some Elementary
Concepts in the Statistical Evaluation of Trace Element
Measurements. Spectrochim. Acta 33B: 625; 1978.
16. Keith, L. H.; Crummett, W; Deegan Jr, J; Libby, R. A.; Taylor,
J. K.; Wentler, G. "Principles of Environmental Analysis",
Analyt. Chem. 1983, 55, 2210.
17. Kateman, G.; Pijpers, F. W. Quality Control in Analytical
Chemistry. New York: John Wiley & Sons; 1981.
18. Heydorn, K.; Wanscher, B. Application of Statistical Methods
to Activation Analytical Results Near the Limit of Detection.
Fresenius' Z. Anal. Chem. 292(1): 34-38; 1978, See also:
Heydorn, K. Neutron Activation Analysis for Clinical Trace
Element Research, 2 Vol. Boca Raton: CRC Press; 1984
19. Feigl, F. Tpfel
Arbeits-methoden, Mikrochemi
20. Kaiser, H. Z. Anal. Chem. 209: 1; 1965 [Ref. 32].
Kaiser, H. Z. Anal. Chem. 216: 80; 1966.
Kaiser, H. Two Papers on the Limit of Detection of a Complete
Analytical Procedure, English translation of the above
manuscripts. London: Hilger; 1968.
21. Altshuler, B.; Pasternack, B. Statistical Measures of the
Lower Limit of Detection of a Radioactivity Counter. Health
Physics 9: 293-298; 1963.
22. Wilson, A. L. The Performance-Characteristics of Analytical
Methods. Talanta 17: 21; 1970; 17; 31: 1970; 20: 725;
1973; and 21: 1109; 1974.
23. Currie, L. A. Limits for Qualitative Detection and
Quantitative Determination. Anal. Chem. 40(3): 586; 1968.
24. St. John, P. A.; Winefordner, J. D. A Statistical Method for
Evaluation of Limiting Detectable Sample Concentrations.
Anal. Chem. 39: 1495-1497; 1967.
25. Nicholson, W. L. "What Can Be Detected"; Developments in
Applied Spectroscopy, v.6, Plenum Press, p. 101-113, 1968;
Nicholson, W. L . , Nucleonics, 24 (1966) 118.
26. Grinzaid, E. L.; Zil'bershtein, Kh. I.; Nadezhina, L. S.;
Yufa, B. Ya. Terms and Methods of Estimating Detection Limits
in Various Analytical Methods. J . Anal. Chem. - USSR 32:
1678; 1977.
27. Ingle, J. D., Jr. Sensitivity and Limit of Detection in
Quantitative Spectrometric Methods. J. Chem. Educ. 51(2):
100-5; 1974.
28. Currie, L. A. "Lower Limit of Detection: Definition and
Elaboration of a Proposed Position for Radiological Effluent
and Environmental Measurements," U S Nuclear Regulatory
Commission, NUREG/CR-4007, 1984.
29. IUPAC Comm. on Spectrochem. and other Optical Procedures for
Analysis. Nomenclature, symbols, units, and their usage in
spectrochemical analysis, Pure Appl. Chem. 45 (1976) 99.
30. IUPAC, Commission V.3, Recommendations for Nomenclature in
Evaluation of Analytical Methods, Draft Report, 1986.
31. Kaiser, H. Spectrochim. Acta, 1947, 3, 40.
32. Kaiser, H. Z. Anal. Chem., 1965, 209, 1.
33. W. B. Crummett in Ref. 71.

34. Brossman, M. W., Kahn, H; King; D.; Kleopfer, R.; McKenna; G.;
Taylor, J . K. Reporting of Low-Level Data for Computerized
Data Bases - Chapt. 17 in this volume.
35. Currie, L. A . ; Parr, R. M. "Perspectives on Detection Limits
for Nuclear Measurements in Selected National (US) and
International (IAEA) Programs" - Chapt. 9 in this volume.
36. Cheeseman, R. V . ; Wilson, A. L. Manual on Analytical Quality-
Control for the Water Industry - Relating to the Concept of
Limit of Detection, WRc Environment, Water Research Center,
Medmenham, UK, 1978.
37. Ramos, L. S.; Beebe, K. R.; Carey, W. P.; Sanchez, M. E.;
Erickson, B. C.; Wilson, B. E.; Wangen, L. E.; Kowalski, B. R.
"Chemometrics"; Anal. Chem., 1986, 58, 294R. [Review and
bibliography].
38. Long, G. L.; Winefordner, J . D. "Limit of Detection: A
closer look at the IUPAC definition"; Anal Chem. 1983; 55;
712A.
39. Natrella, M. G. 'Th
and Tests of Significance'; in Ku, H, Ed. "NBS Spec Publ 300";
1969.
40. Smit, H. C.; Steigstra, H. "Noise and Detection Limits in
Signal Integrating Analytical Methods"; - Chapt. 7 in this
volume.
41. Epstein, M. S. "Comparison of Detection Limits in Atomic
Spectroscopic Methods of Analysis"; - Chapt. 6 in this volume.
42. Ku, H. H. Edit., Precision Measurement and Calibration, NBS
Spec. Public. 300 (1969), p. 315.
43. Saw, J . G.; Yang, M. C. K.; Mo, T. C. 'Chebyshev Inequality
with Estimated Mean and Variance'; The Amer. Statistician;
1984; 38; 130.
44. Johnson, J . E.; Johnson J . A. "Radioactivity Analyses and
Detection Limit Problems of Environmental Surveillance at a
Gas-Cooled Reactor" - Chapt. 14 in this volume.
45. Snedecor, G. W.; Cochran, W. G. Statistical Methods, 6th
Edit., Iowa State Univ. Press. (1973).
46. Kingston, H. M.; Greenberg, R. R.; Beary, E. S.; Hardas, B.
R.; Moody, J . R.; Rains, T. C.; Liggett, W. S. "The
Characterization of the Chesapeake Bay: A Systematic Analysis
of Toxic Trace Elements"; National Bureau of Standards,
Washington, DC, 1983, NBSIR 83-2698.
47. Scales, B. Anal. Biochem. 5, 489-496 (1963).
48. Currie, L. A. "Model Uncertainty and Bias in the Evaluation
of Nuclear Spectra"; J . of Radioanalytical Chemistry 39, 223-
237 (1977).
49. Koch, W.; Liggett, W. "Critical Assessment of Detection
Limits for Ion Chromatography" - Chapt. 11 in this volume.
50. Watters, R. L.; Wood, L. J. in Ref. 71.
51. Murphy, T. J . The Role of the Analytical Blank in Accurate
Trace Analysis, NBS Spec. Publ. 422, Vol. I I , U. S. Government
Printing Office, Washington, DC, 509 (1976).
52. Kelly, W. R.; Hotes, S. A. "The Importance of Chemical Blanks
and Chemical Yields in Accurate Chemical Analysis"; Preprint
(1987).

53. Kelly, W. R.; Fassett, J . D.; Hotes, S. A., 'Determining

Picogram Quantities of U in Human Urine by Thermal Ionization
Mass Spectrometry'; Health Physics, 1987; 52, 331.
54. Currie, L. A. 'Quality of Analytical Results, with Special
Reference to Trace Analysis and Sociochemical Problems'; Pure
& Appl. Chem., 1982, 54, 715.
55. Currie, L. A. The Limit of Precision in Nuclear and
Analytical Chemistry. Nucl. Instr. Meth. 100: 387; 1972.
56. Lub, T. T.; Smit, H. C., Anal. Chim. Acta, 1979, 112, 341.
57. Mellor, R. A.; Harrington, C. L. "Evaluating the Impact of
Hypothesis Testing on Radioactivity Measurement Programs at a
Nuclear Power Facility" - Chapt. 13 in this volume.
58. Parr, R. M.; Houtermans, H.; Schaerf, K. 'The IAEA
Intercomparison of Methods for Processing Ge(Li) Gamma-Ray
Spectra'; in "Computers in Activation Analysis and Gamma-Ray
Spectroscopy"; U.S Dept of Energy Sympos Ser 49 1979
544.
59. Liggett, W. ASTM Conf
Measurements, 1984, Boulder, CO.
60. Watters, R. L . ; Spiegelman, C. H., and Carroll, R. J .
"Heteroscedastic Calibration in Inductively Coupled Plasma
Spectrometry"; Anal. Chem., 1987, 59, 1639.
61. Currie, L. A. "The Many Dimensions of Detection in Chemical
Analysis"; in Chemometrics in Pesticide/Environmental Residue
Analytical Determinations, ACS Sympos. Series (1984).
62. Owens, K. G.; Bauer, C. F.; Grant, C. L. "Effects of
Analytical Calibration Models on Detection Limit Estimates";-
Chapt. 10 in this volume.
63. Iyengar, G. V.; Tanner, J . T.; Wolf, W. R.; Zeisler, R.
'Preparation of a Mixed Human Diet Material for the
Determination of Nutrient Elements, Selected Toxic Elements
and Organic Nutrients: a Preliminary Report', submitted to
The Science of the Total Environment, 1986.
64. Currie, L. A. Detection and Quantitation in X-ray
Fluorescence Spectrometry, Chapter 25, X-ray Fluorescence
Analysis of Environmental Samples, T. Dzubay, Ed., Ann Arbor
Science Publishers, Inc., p. 289-305 (1977).
65. Horwitz, W.; in Ref. 71.
66. Zweig, M. "Establishing Clinical Detection Limits of
Laboratory Tests"; - Chapt. 8 in this volume.
67. Natrella, M. G.; Experimental Statistics, NBS Handbook 91,
1963.
68. Dixon, W. J.; Massey, F. J. Introduction to Statistical
Analysis, McGraw-Hill, New York, 1957.
69. Schlain, B., personal communication, 1987.
70. Currie, L. A. 'The Limitations of Models and Measurements as
Revealed through Chemometric Intercomparison'; J . Res. NBS,
1985, 90, 409.
71. Kurtz, D.; Taylor, J . K.; Sturdivan, L . ; Crummett, W.;
Midkiff, C.; Watters, R.; Wood, L . ; Hanneman, W.; Horwitz, W.
Real-World Limitations to Detection: A Panel Discussion;
Chapt. 16 in this volume.
72. Currie, L. A.; Gerlach, R. W.; Lewis, C. W.; Balfour, W. D.;
Cooper, J . A.; Dattner, S. L.; DeCesar, R. T.; Gordon, G. E.;
Heisler, S. L.; Hopke, R. K.; Shah, J . J.; Thurston, G. D.;

Williamson, H. J. "Interlaboratory Comparison of Source

Apportionment Procedures: Results for Simulated Data Sets";
Atmospheric Environment 18 (1984), 1517.
73. Eisenhart, C.; Zelen, M. Ch 12, "Elements of Probability"; in
Handbook of Physics, E. U. Condon and H. Odishaw, Ed., McGraw-
Hill, New York, 1958.
74. Kendall, M.; Stuart, A.; Ord, J . K. The Advanced Theory of
Statistics; Vol. 3, MacMillan; New York,; 1983.
75. Delaney, M. F. 'Multivariate Detection Limits for Selected
Ion Monitoring Gas Chromatography - Mass Spectrometry';
Chemometrics and Intelligent Laboratory Systems, 1987.
76. Wold, S.; Albano, C.; Dunn, III, W. J.; Edlund, U.; Esbensen,
K.; Geladi, P.; Hellberg, S.; Johansson, E.; Lindberg, W.;
Sjstrm, M. "Multivariate Data Analysis in Chemistry"; in
Chemometrics: Mathematics and Statistics; B. R. Kowalski,
Ed., (Reidel Publishing Co.) 1984; pp 17-96
77. Draper, N.; Smith
York, 1981.
78. McCormack, M. "Realistic Detection Limits and the Political
World" - Chapt. 2 in this volume.
79. Slovic, P. "Perception of Risk"; 280-285 in Ref. 4.
80. Lave, L. "Health and Safety Risk Analysis: Information for
Better Decisions"; 291-295 in Ref. 4.
81. Handler, P. Dedication Address, Northwestern Univ. Cancer
Center, 1979. (See also S. J. Gould in the New York Times
Magazine, 19 April 1987.)
82. Nieboer, E.; Jusys, A. A. "Contamination Control in Routine
Ultratrace Analysis of Toxic Metals"; Ch. 1 in Chemical
Toxicology and Clinical Chemistry of Metals, S. S. Brown and
J. Savory, Eds., Academic Press, London, 1983.
83. Sharaf, M. A.; Illman, D. L.; Kowalski, B. R. Chemometrics,
Wiley, New York, 1986.
84. Kowalski, B. R., Ed. Chemometrics: Mathematics and
Statistics in Chemistry, (Reidel Publishing Co.) 1984.
85. Hurst, G. S.; Payne, M. G.; Kramer, S. D.; Young, J. P. Rev.
Mod. Phys., 1979, 51, 767.
86. Donaldson, W. T., Envir. Sci. Tech., 1977, 11, 348.
87. Guinn, V., personal communication; 1987.
88. Donahue, D., Fourth Intern. Sympos. on Accelerator Mass
Spectrometry, Niagara on the Lake, 1987.
89. Heydorn, K.; Christensen, L. H. "Verification Testing of
Commercially Available Computer Programs for Photpeak Area
Evaluation"; Intern. Conf. on Meth. Appl. Radioanalytical
Chemistry, Kona, 1987.
90. Cochran, W. G.; Biometrics, 1964, 20, 191.
91. Andrews, R. M. "Meat Inspector: 'Eat at Own Risk'"; Wash.
Post, May 16, 1987. [Science news following issuance of a
National Academy of Sciences Report concerning food poisoning
from undetected microorganisms.]
92. Lochamy, J . D. The minimum-detectable-activity concept. Nat.
Bur. Stand. (U.S.) Spec. Publ. 456; 1976, 169-172.
93. IAEA; Users' Guide on Limit of Detection; in preparation.
(See Ref. 35).

94. "Standard Radiological Effluent Technical Specifications for

Pressurized Water Reactors," U. S. Nuclear Regulatory
Commission, NUREG-0472, Rev. 3, September 1982.
95. Federal Register, 1984, 49, 43431.
96. Federal Register, 1984, 49, 43430; and Glaser, J., Foerst, D.,
McKee, G., Quave, S., and Budde, W., "Trace Analysis for
Wastewaters," Environ. Sci. Tech., 1981, 15, 1426.
97. Forina, M.; Lanteri, S. "Data Analysis in Food Chemistry;"
pp. 305-349; in Ref. 84.
98. Feigl, F. Chemistry of Specific, Selective and Sensitive
Reactions; New York: Academic Press; 1949.
99. Emich, F. Ber. 1910; 43; 10.
100. Cox, D. R.; Lewis, P. A. W. The Statistical Analysis of
Series of Events; London: Methuen; 1966.
Notes
Note 1. Analytica possibility
"single atom detection" (85) . At the same time i t is recognized
15
that at concentrations of 1 part in 10 (in water) in principle
"every known organic compound could be detected" (86) These
measurement realities mandate the setting of regulatory levels on
bases other than either non-zero concentrations, or the inherent
a b i l i t y to detect.
Note 2. That Feigl's "Identification Limit" referred to the
minimum quantity detectable (Lp) as opposed to the decision or
c r i t i c a l level (L^) is clear from his statement defining the
"'Erfassungsgrenze' [as] die Kleinste absolute Menge Substanz ...
die ... noch nachweisbar und bestimmbar ist "(Ref. 19, p. 6). In a
later, english language publication, this meaning was amplified in
a manner that foreshadowed the modern s t a t i s t i c a l approach to
detection. In the volume "Chemistry of Specific, Selective, and
Sensitive Reactions", p. 14 (8), Feigl described a test for
magnesium which was "always" positive, for 40 repetitions, using a
0.05% Mg solution. With dilution by factors of 10 and 50, however,
the test was positive only in 24 and 6 instances, respectively.
With this, Feigl embraced the concept of the "region of uncertain
reaction" (99), and a condition for the identification limit that
the chance of a false negative be negligible.
Note 3. Symbols introduced in this section include the
following: y [gross signal], B [null signal = background, baseline,
or blank], S [net signal], x [analyte concentration or amount], A
["sensitivity" or calibration factor], pdf [probability density A
function], cdf [cumulative distribution function], superscript or 2
est( ) [estimated value] , E() or /i [expected value] , V or cr
2
[population variance], s [estimated variance], o [standard Q
deviation of the estimated net signal, when E(S)=0], CI [confidence

interval], df [degrees of freedom], A [bias], A [bias detection D
limit; discrimination l i m i t ] . Subscripts, , , denote lower and

m M
upper limits, respectively.

Note 4 . I n a r e c e n t paper ( 4 3 ) , Saw, Yang, and Mo p r e s e n t an

e x t e n s i o n o f the Chebyshev I n e q u a l i t y t h a t uses e s t i m a t e d v a l u e s
f o r the mean and v a r i a n c e . As e x p e c t e d , the l i m i t s are g e n e r a l l y
b r o a d e r ( o r l i m i t i n g a's l a r g e r ) t h a n w o u l d be the case w i t h o-
known, b u t the extended i n e q u a l i t y i s d i r e c t l y a p p l i c a b l e t o our
t e s t o f the n u l l h y p o t h e s i s i n the form: B - B, where the mean
v a l u e f o r the b l a n k B i s based on n - r e p l i c a t e s . An important
c o n c l u s i o n t h a t can be d e r i v e d from the t r e a t m e n t o f Saw, e t a l . i s
t h a t the s m a l l e s t l i m i t i n g v a l u e f o r a t h a t can be o b t a i n e d w i t h n-
r e p l i c a t e s i s l / ( n + l ) . T h i s v a l u e i s o b t a i n e d once k i s a t l e a s t
7(n+l). Thus, f o r 5 r e p l i c a t e s , one can do no b e t t e r t h a n a
( l i m i t ) = 0 . 1 6 7 , b u t the r e q u i s i t e s t a n d a r d d e v i a t i o n m u l t i p l i e r k
need be no l a r g e r t h a n JZ w h i c h e q u a l s 2 . 4 5 . L a r g e r k's would
b r i n g no improvement. To a t t a i n an upper l i m i t o f 0 . 0 5 f o r the
f a l s e p o s i t i v e r i s k , n must be a t l e a s t 1 9 . The corresponding
v a l u e f o r k w o u l d be 7 2 0 o r 4 . 4 7 The c r i t i c a l l e v e l S f o r the
d e t e c t i o n d e c i s i o n woul
estimated standard d e v i a t i o
freedom. S i n c e a(B-B) - a B 7(n+l)/n, S = 4 . 4 7 7 ( 2 0 / 1 9 ) s , or
c B
4.59 s . B Had we been a b l e t o assume n o r m a l i t y , the m u l t i p l i e r 4 . 4 7

w o u l d have been r e p l a c e d w i t h 1 . 7 3 , Student's t f o r a 1 - s i d e d t e s t
a t the 0 . 0 5 s i g n i f i c a n c e l e v e l .
Note 5. The r e l a t i v e s t a n d a r d d e v i a t i o n [RSD] o f a b a s e d on
o b s e r v e d counts i s 1/(27/*) f o r P o i s s o n d a t a , o r about 6% f o r

/i=E(N)=60. E q u i v a l e n t p r e c i s i o n f o r a b a s e d on r e p l i c a t i o n would
i
r e q u i r e about 2/i o r 120 degrees o f freedom. The same i s t r u e f o r

c o n f i d e n c e i n t e r v a l s f o r / i , hence V , b a s e d on c o u n t s , vs V b a s e d

2 t o
on r e p l i c a t i o n . For more d e t a i l , i n c l u d i n g the use o f x derive
b o t h t y p e s ( c o u n t s , r e p l i c a t i o n ) o f CI's see Ref. 68 and the
monograph by Cox and Lewis ( 1 0 0 ) . Adequacy o f the l a r g e count
(normal) a p p r o x i m a t i o n , and the e x a c t t r e a t m e n t f o r extreme low-
l e v e l d a t a (/x 10 o r l e s s ) are covered i n Ref. 2J3 and the
references therein.
Note 6. Combining the i n e q u a l i t y c o n s t r a i n t [V x > 0] w i t h

known i n t e r n a l P o i s s o n v a r i a n c e has been u s e d by G u i n n f o r t o t a l
v a r i a n c e e s t i m a t i o n i n a c t i v a t i o n a n a l y s i s ( 8 7 ) , by Donahue (88)
f o r the same purpose i n a c c e l e r a t o r mass s p e c t r o m e t r y , by Heydorn
(18.89) f o r the a n a l y s i s o f p r e c i s i o n o f gamma r a y spectrum
a n a l y s i s , and by C u r r i e (55) f o r optimal weighting i n counting
experiments. The v a r i a n c e w e i g h t e d t a p p r o x i m a t i o n d e r i v e s from
work by Cochran ( 9 0 ) . Cochran's work, however, a p p l i e d t o a
somewhat d i f f e r e n t model t h a n used h e r e . The s t a t i s t i c a l p r o p e r -
t i e s o f t h i s e x t e n s i o n o f h i s t e c h n i q u e have n o t been s t u d i e d .
Note 7 . M e t h o d - s p e c i f i c m e c h a n i s t i c u n d e r s t a n d i n g i s , i n the
l a s t a n a l y s i s , the o n l y r o u t e t o r e l i a b l e measurements, r e l i a b l e
detection limits, and m e a n i n g f u l societal decisions involving
science related public policy. The need f o r u n d e r s t a n d i n g on the
m e c h a n i s t i c l e v e l becomes even c l e a r e r when one c o n s i d e r s the
i n h e r e n t l i m i t a t i o n s o f d e c i s i o n s and p o l i c i e s b a s e d s o l e l y on the
empirical record, such as certain aspects of epidemiology.
S a m p l i n g s t a t i s t i c s and c o n t r o l o f the system under i n v e s t i g a t i o n
e m p i r i c a l l y , c o n s t i t u t e severe l i m i t a t i o n s -- ones t h a t can be

overcome only by d i r e c t i n g our research p r i o r i t i e s toward increased

basic s c i e n t i f i c knowledge (81).
Note 8. The drawback of p o l i c i e s based on "detection l i m i t s " so

defined that - 1^, , where the false negative r i s k () i s 50%, and
the f a l s e p o s i t i v e (a) i s 10 to 400 times smaller, i s s t r i k i n g i n
the l i g h t of matters a f f e c t i n g large sectors of the public. A case
i n point: food poisoning a f f e c t s several m i l l i o n c i t i z e n s i n the US
each year "from poultry because federal inspection f a i l s to detect
contamination by salmonella and other b a c t e r i a l microorganisms"
(91). In situations l i k e this we must ask how many would be
content with a test whose "detection l i m i t " i s designed to catch
only h a l f the contaminated specimens! In comparison, one wonders
how society i s better served i n such an instance by reducing false
positives to a n e g l i g i b l e l e v e l . (See also sections 2, 3.2.1,
3.3.1, and F i g . 10.)
R E C E I V E D September 22, 198

Chapter 2
Realistic Detection Limits and the Political World
Mike McCormack
McCormack Associates, Inc., 508 A Street, S.E., Washington, DC 20003
My comments in this chapter are not particularly technical,

unless one considers the subject of constructive and realistic
political activity to be "social engineering". I propose
to explore the interface between the realities of detection
limits on the one hand; and the political world, and our
obligations within i t , on the other.
Probably the best starting point for such a discussion
is the recognition that we are all part of the political
community. Responsible scientists cannot avoid the obligations
of being citizen-politicians in any free society. Scientists
are a special subset of that greater political community,
and we have, over recent decades, attracted ever-increasing
attention within it. That attention has constituted a blessing
in times of success, with acclaim and generous financial
support; and a frustration in times of public confusion,
incomprehension and fear.
The people of any society may, of course, always be categor-
ized according to any number of criteria. For the purpose
of this discussion, it is fair, and not uncomplimentary,
to define two groups within the population of this country: one
in which the concepts of very large and very small numbers
are comprehended, and operations with them easily understood,
and in which the scientific method and especially the
meaningful interpretation of analytical data is appreciated;
and the other group, within which this is not generally the
case. For simplicity, we usually refer to these two groups
as the scientific, and the non-scientific communities.
0097-6156/88/0361 -0064$06.00/0
1988 American Chemical Society

2. McCORMACK Realistic Detection Limits and the Political World 65
I t frequently appears to me that there i s a mutuality

of ignorance (and perhaps a "love-hate" relationship) binding
and at the same time d i v i d i n g these two groups. For
instance, a n a l y t i c a l chemists frequently seem unable to understand
that their f e l l o w - c i t i z e n s without technical t r a i n i n g cannot
understand what the chemists, or many other s c i e n t i s t s and
mathematicians, are t a l k i n g about when they attempt to convey
s c i e n t i f i c or mathematical information to the public on matters
of s o c i e t a l concern, or why their information and recommendations
are not r e a d i l y accepted. After a l l , this attitude assumes,
s c i e n t i s t s know what they're talking about.
Equally confusing, i n the minds o f many members of the
non-scientific l a y p u b l i c , i s the generally more objective,
l e s s emotional a t t i t u d e that most s c i e n t i s t s take toward
presumed threats to human health and safety that the public
i s told flow fro
e s p e c i a l l y i n the
world. "Those guys don't care they're p r o s t i t u t i n g their
i n t e g r i t y for a paycheck," so that attitude concludes.
The ancient enemies o f humankind have always been ignorance
and fear, which i n turn are the progenitors of s u p e r s t i t i o n ,
hatred, bigotry and the i s o l a t i o n o f one group from another.
To a limited but disturbing extent, t h i s phenomenon has developed
within t h i s country; and today a g u l f o f confusion, suspicion,
fear, emotionalism and sometimes h o s t i l i t y l i e s between s i g -
n i f i c a n t portions of the s c i e n t i f i c and n o n - s c i e n t i f i c commun-
ities .
Within the l e g i s l a t i v e arena, and within some portions
of the press and e l e c t r o n i c media, t h i s has focused on r e a l
or presumed r i s k s to human health and safety from the presence,
or the possible presence, of " a r t i f i c i a l " (that i s , d e l i b e r a t e l y
manufactured by humans) chemicals i n our food, water, medicines,
a i r , the things we touch, or the ground upon which we walk.
This often leads to the enactment of u n r e a l i s t i c l e g i s l a t i o n ,
and to regulations that attempt to override the basic laws
of nature and the r e a l i t i e s o f the physical world.
The challenge that we i n the s c i e n t i f i c community face
arises from the fact that frequently only we understand that
such a contradiction e x i s t s , l e t alone what a r a t i o n a l solution
to i t may be. This s i t u a t i o n i s further complicated by the
f a c t that the fear we are trying to overcome often gives
r i s e to an emotionalism within which s c i e n t i s t s and even
s c i e n t i f i c f a c t become suspect i n the minds o f some
non-scientists. I t i s not uncommon to encounter the argument
that s c i e n t i s t s i n a s p e c i f i c d i s c i p l i n e should be d i s q u a l i f i e d
from providing information on issues involving that d i s c i p l i n e
because "they are prejudiced".
In addition, a s i g n i f i c a n t number of i n d i v i d u a l s , from
the press and electronic media, from c e r t a i n a c t i v i s t and
r e l i g i o u s groups, and even among elected public o f f i c i a l s ,
often encourage emotionalism with respect to science i n the
minds of susceptible i n d i v i d u a l s . At that point, s c i e n t i s t s
may throw up t h e i r hands, and tend toward open r i d i c u l e o f ,
and h o s t i l i t y toward, such i n d i v i d u a l s .

Of c o u r s e , t h i s only exacerbates the s i t u a t i o n . Since

we are generally the ones who understand the basics of science
related issues, i t i s incumbent upon us, as responsible c i t i z e n s
i n a f r e e s o c i e t y , to speak out, p a t i e n t l y helping our
non-scientific fellow c i t i z e n s understand; and leading them
to responsible, constructive perspectives. This i s a never
ending o b l i g a t i o n .
Before anyone i n the s c i e n t i f i c community (and e s p e c i a l l y
i n t h i s book) becomes over-inflated with self-righteousness,
l e t me emphasize that we, as individuals and as a profession,
are f a r from f a u l t l e s s or free from legitimate c r i t i c i s m
with respect to our r e s p o n s i b i l i t i e s . Certainly more members
o f the s c i e n t i f i c community and chemists i n p a r t i c u l a r
should have recognized at an e a r l i e r date the hazards
of i n d u s t r i a l p o l l u t i o n or the unlimited application of c e r t a i n
p e s t i c i d e s to the environment i which liv
and work and play.
Today i t seems
could have been so unconcerned only several decades ago that
he o r she would not speak up to protest such p r a c t i c e s .
In addition, many members of the s c i e n t i f i c community have
seriously damaged meaningful communication with the non-scientific
community by the display o f an unbecoming arrogance; f a i l i n g
to appreciate that t h i s increases h o s t i l i t y among those already
suspicious.
These comments are not intended as c r i t i c i s m or soul
searching, but to help put the s i t u a t i o n into honest context.
The fact i s that there i s a broad chasm of ignorance, suspicion
and fear that separates the s c i e n t i f i c community from much
of the p u b l i c , the press and l e g i s l a t o r s at the state and
federal l e v e l .
An increasing volume of l e g i s l a t i o n , s i n c e r e l y intended
to p r o t e c t the p h y s i c a l environment, and to insure human
h e a l t h and safety has been, and i s being enacted i n t h i s
atmosphere. Much of the l e g i s l a t i o n has been proposed as
a r e s u l t of honest concerns because of accidents involving
chemicals or from emotional reaction to suggestions that
some chemical which humans may contact may be a carcinogen
or cause other physical harm. L e g i s l a t i o n i s often written
i n ignorance of the actual degree of hazard involved; and
allowable concentrations of hazardous substances have frequently
been e s t a b l i s h e d without an adequate understanding of the
r e s u l t i n g impact.
This i s a s i g n i f i c a n t problem f o r a l l society, and i t
w i l l probably continue u n t i l vigorous i n i t i a t i v e s are undertaken
from within the s c i e n t i f i c community to a s s i s t more public
o f f i c i a l s and more members of the news media appreciate the
r e a l i t i e s involved and the importance of observing them.
To be f a i r , the concepts involved i n such l e g i s l a t i o n
make good sense to a casual observer from outside the s c i e n t i f i c
community. I t seems obvious and l o g i c a l to many honest and
sincere c i t i z e n s that i f a substance i s " t o x i c , " "hazardous"
or "carcinogenic" that none of i t should be allowed i n our
food or drink, i n the a i r we breathe or the ground on which

McCORMACK Realistic Detection Limits and the Political World
we walk; none at a l l . The concentration, we are t o l d , should

be zero. "Why not?"
The suggestion that i t i s not p r a c t i c a l to t r y to reduce
the concentration o f a hazardous substance to the minimal
l e v e l at which i t can be detected (even ignoring cost) frequently
seems t o c o n s t i t u t e a "cop-out"; and those who make such
a suggestion may be suspected o f being i n league with some
deliberate p o l l u t e r who would c a l l o u s l y endanger the health
(or the l i v e s ) o f our c h i l d r e n .
Analytical techniques that have been developed can now,
for most substances, detect concentrations far below hazardous
l e v e l s . The effect of these advances i n a n a l y t i c a l capability
should be cause for s a t i s f a c t i o n . They w i l l help make pos-
s i b l e future research, both theoretical and applied, including
studies r e l a t i v e to human health and safety.
However, there ar
that would provide tha ( )
designated as carcinogens should be "zero", or (2) that the
concentration o f some contaminants i n drinking water should
be as low as detectable. Thus, each new accomplishment i n
a n a l y t i c a l technologies that pushes the l i m i t o f detection
o f any suspect chemical to a lower concentration, brings
with i t the potential f o r new p o l i t i c a l problems.
Such u n r e a l i s t i c provisions i n the law or regulations
are, i n r e a l i t y , s e l f - d e f e a t i n g . However, i t i s d i f f i c u l t
to persuade the non-scientist that they do not protect the
public, and frequently cause the waste o f a great deal o f
public and private money.
There are ways to help the average c i t i z e n understand
t h i s . For instance, at one part per b i l l i o n , there are s t i l l
t r i l l i o n s o f molecules o f a foreign substance i n a glass
o f water. The average c i t i z e n i s shocked to learn t h i s .
Another way i s to point out that i n the r e a l world of drinking
water and food, at parts per b i l l i o n , there i s some amount
of almost everything i n almost everything. Another way to
express t h i s t r u t h i s that a "chemically pure" substance
(99-9999$ pure) s t i l l contains one part per m i l l i o n impurities.
This i s equivalent to 100 d i f f e r e n t substances at 10 parts
per b i l l i o n ; or, i f you l i k e , 10 foreign substances at 100
parts per b i l l i o n . Thus i t becomes extremely d i f f i c u l t to
analyze a c c u r a t e l y f o r many substances at the l e v e l o f a
few p a r t s p e r b i l l i o n because o f p o t e n t i a l contamination
of the system from equipment, and from reagent impurities.
(Most n o n - s c i e n t i s t s do not appreciate that a b i l l i o n
i s one thousand m i l l i o n and do not comprehend what a t r i l l i o n
or one part per t r i l l i o n means. Here i s an i l l u s t r a t i o n
that may be o f value. Imagine ordinary glass marbles about
1/2 inch i n diameter similar to the ones we played with
when we were k i d s . An ordinary square card table w i l l hold
about 10,000 such marbles, one layer thick, packed as densely
as possible. For comparison, i t would take about two m i l l i o n
marbles to cover the f l o o r o f an average size lecture h a l l .

It would take about 40 b i l l i o n marbles to cover Central Park

i n New York C i t y , one layer deep. Thus, one part per b i l l i o n
would be about 40 marbles among a layer of marbles completely
covering Central Park.
We can r o u t i n e l y detect 40 parts per b i l l i o n of many
substances i n water, but not necessarily i n natural systems
containing other unknown substances i n higher concentrations;
and i t may not be p r a c t i c a l , or even possible, to guarantee
the removal of those 40 parts per b i l l i o n from any natural
system without e x o r b i t a n t cost. One t r i l l i o n marbles
one thousand b i l l i o n , or a m i l l i o n m i l l i o n , would cover
a l l o f Manhattan Island and a l l of the Hudson River from
the George Washington Bridge down to the Battery. Even i n
those few cases when we can detect a few parts per t r i l l i o n
the equivalent of a handful of marbles scattered somewhere
over Manhattan Islan
would probably not mak
You must help mak point,
and standards s e t t i n g are separate a c t i v i t i e s , the one comprising
a s t r i c t l y technical a c t i v i t y and the other, a s o c i e t a l value
judgment. The b a s i c r a t i o n a l e for disengaging these two
a c t i v i t i e s i s that the ubiquitous presence of other carcinogenic
agents a l l around us swamps the p o t e n t i a l hazard, even
of known carcinogens, at extremely low concentrations.
Unfortunately, there has been too much uncertainty and
confusion regarding the r e a l meaning and r e l i a b i l i t y of detection
l i m i t s f o r t h e i r optimal use i n the world of l e g i s l a t i o n
and regulation. I urge you to work toward a consistent d e f i n i t i o n
and meaningful r e a l i z a t i o n of the l i m i t of detection from
a t e c h n i c a l point of view, so that i t can be accepted by
society as a l i m i t above which chemicals w i l l be found ( i f
present) with an acceptable degree of c e r t a i n t y . This means
that such l i m i t s must f u l l y take into account the r e a l sample
c h a r a c t e r i s t i c s including the e f f e c t s of interference, contam-
i n a t i o n , e t c . Otherwise, spurious l e v e l s f o r f a l s e positives
or f a l s e negatives may lead to regulations that are too r e -
s t r i c t i v e or to pressures that make them too l e n i e n t .
As with a l l s c i e n t i f i c measurements, working at l i m i t s
o f d e t e c t i o n n e c e s s a r i l y involves a n a l y t i c a l uncertainty.
However, t h i s c r e a t e s a s p e c i a l problem i n communication
where such uncertainties are not generally understood. We
must assure that our detection l i m i t s and, i n f a c t , a l l of
our a n a l y t i c a l measurements are established with an adequate
l e v e l of " c e r t a i n t y " , e s p e c i a l l y as assurance for p o l i c y
makers seeking responsible standards. After a l l , they must
assure t h e i r c o l l e a g u e s , constituents and the press that
standards based on r i s k assessment make the most sense, and
that s c i e n t i f i c a l l y v a l i d data have been used i n making such
assessments.
F i n a l l y , detecting a minute trace of a contaminant does
not n e c e s s a r i l y make i t s economic removal p r a c t i c a l . This
i s not immediately obvious to the average layman, but i t
can be e a s i l y explained by extrapolating the concept to the
point of absurdity. Can we remove one part per million?

McCORMACK Realistic Detection Limits and the Political World
I f so, i s i t s t i l l possible to remove one part per b i l l i o n ?

I f s o , and we find one part per t r i l l i o n , i s i t possible
to remove i t , etc.?
These r e a l i t i e s must be borne i n mind by s c i e n t i s t s who
work i n the realm of detection l i m i t s when reporting their
findings to non-scientists, and e s p e c i a l l y to lawmakers.
This brings me back to our o b l i g a t i o n as c i t i z e n s . I
hope you w i l l take this matter s e r i o u s l y .
We a l l f a c e a monumental challenge. Fortunately, we
have s c i e n t i f i c truth and common sense as our most e f f e c t i v e
tools.
We must add to that a dedication to serving our fellow
c i t i z e n s by helping them understand, and putting the technology
and i t s purposes into proper context.
R E C E I V E D September 22, 1987

Chapter 3
Scientific Measurements
and Data
in Public Policy-Making
Thomas H. Moss
Case Western Reserve University, Cleveland, OH 44106
Sound technica
data must seem
irrelevant to the course of public policy
decision-making. Recent experience in the politics
of clean air, acid rain, toxic substances control,
pesticide regulatory legislation, as well as other
debates, indicates a more tempered view, however.
On closer analysis the role of scientific measurement
is seen as a vital one, but one which requires a
sense of timing and perception of the dynamics of
human behavior in seeking solutions to challenging
problems.
Technical measurements and t h e r e s u l t i n g d a t a a r e used i n two
d i s t i n c t ways i n t h e contemporary American p o l i c y d e c i s i o n - m a k i n g
process. On t h e one hand i s t h e i r r o l e i n t h e s y s t e m a t i c b u i l d i n g
of a body o f knowledge w h i c h i d e a l l y becomes t h e b a s i s f o r p u b l i c
policy. On t h e o t h e r hand i s t h e i r use as weapons i n a war o f
words, o r c o n t e s t f o r public attention and s u p p o r t . At times
one or the other of these functions may seem t o dominate, b u t
i t i s my t h e s i s i n t h i s e s s a y t h a t i t i s e x t r e m e l y r i s k y f o r a l l
concerned t o n e g l e c t e i t h e r aspect. Beyond t h a t , my own e x p e r i e n c e
t e l l s me t h a t p a r t i c i p a n t s i n p u b l i c p o l i c y d i s c u s s i o n s can, t o
a c o n s i d e r a b l e degree, create t h e i r own r e a l i t y from t h e c h o i c e
between these two extremes. I t i s the r e a l i z a t i o n of t h i s
r e l a t i v e l y strong degree o f p e r s o n a l c o n t r o l o v e r t h e n a t u r e o f
the debate which i s so i m p o r t a n t t o p a r t i c i p a n t s . By e x e r t i n g
that control wisely, p a r t i c i p a n t s c a n c r e a t e a p r o c e s s i n which
t h e y have c o n f i d e n c e ; by u s i n g i t unwisely, i n c o n s i s t e n t l y , or
l a z i l y , they can f i n d t h e m s e l v e s i n a p r o c e s s i n w h i c h t h e y w i l l
f e e l v i c t i m i z e d , f r u s t r a t e d , and h e l p l e s s .
Case H i s t o r i e s of Measurement and S c i e n t i f i c Data i n P o l i t i c a l

Decision-Making:
What a r e examples t o j u s t i f y t h i s i n t e r p r e t a t i o n of the regulatory
0097-6156/88/0361 -0070S06.00/0
1 9 8 8 American Chemical Society

3. MOSS Scientific Measurements and Data in Public Policy-Making 71
climate? C l e a r l y my own w i l l be based on i n t e r p r e t a t i o n s o f h i s t o r y

and m o t i v a t i o n s w i t h which some may disagree. However i n t h e
c a s e s I mention below I c o u l d c l e a r l y see p o i n t s a t w h i c h one
p a t h i n the c o u r s e o f p u b l i c p o l i c y debate become dominant o v e r
s e v e r a l p e r f e c t l y f e a s i b l e a l t e r n a t i v e s , o r i n which t h e debate
a b r u p t l y changed c o u r s e . Whether my own i n t e r p r e t a t i o n i s c o r r e c t
o r n o t , my aim i s t o a t l e a s t c h a l l e n g e t h e r e a d e r t o come up
w i t h an a l t e r n a t i v e i n t e r p r e t a t i o n t o h e l p r a t i o n a l i z e t h e c u r i o u s
t w i s t s o f t h i n k i n g on p u b l i c p o l i c y .
The c l a s s i c example i s perhaps t h e l o n g - r u n n i n g debate o v e r
t h e f l a g s h i p o f U.S. e n v i r o n m e n t a l law: the Clean A i r A c t . In
a p i o n e e r i n g 1970 e f f o r t t o e s t a b l i s h and e n f o r c e m e a n i n g f u l l i m i t s
on a i r p o l l u t i o n , t h e f r a m e r s o f the l e g i s l a t i o n chose human h e a l t h
as t h e key and absolute c r i t e r i a f o r enforcement of standards
(1):
"In the Committe
was expressed regardin
technical feasibility
standards. The Committee d e t e r m i n e d t h a t 1) the h e a l t h
o f p e o p l e i s more i m p o r t a n t t h a n t h e q u e s t i o n of whether
t h e e a r l y achievement o f ambient a i r q u a l i t y s t a n d a r d s
p r o t e c t i v e of h e a l t h i s t e c h n i c a l l y f e a s i b l e : and 2)
t h e growth o f p o l l u t i o n l o a d i n many a r e a s , even w i t h
the application of available t e c h n o l o g y , would still
be d e l e t e r i o u s t o p u b l i c h e a l t h .
T h e r e f o r e , t h e Committee d e t e r m i n e d t h a t e x i s t i n g s o u r c e s
o f p o l l u t a n t s e i t h e r s h o u l d meet t h e s t a n d a r d s o f t h e
law o r be c l o s e d down...."
There were many u n d e r s t a n d a b l e reasons f o r the c h o i c e of
t h i s r h e t o r i c at that time. Concern f o r human s u f f e r i n g seemed
t o dominate p u b l i c attitudes, and non-human e c o l o g i c a l effects
seemed much more s u b t l e and distant. Several well publicized
urban a i r p o l l u t i o n e p i s o d e s i n Europe and t h e U.S. had vividly
shown t h e c o n n e c t i o n of a c u t e a i r p o l l u t i o n and human h e a l t h impacts
t h r o u g h t h e d r a m a t i c i n c r e a s e s i n h o s p i t a l a d m i s s i o n s and deaths
from a v a r i e t y of c a r d i a c and pulmonary c a u s e s . On any less
a b s o l u t e s c a l e , however, l e g i s l a t i v e l e a d e r s were w e l l aware t h a t
air q u a l i t y m o n i t o r i n g and e p i d e m i o l o g i c a l d a t a were h o p e l e s s l y
i n a d e q u a t e t o demonstrate the s u b t l e t i e s o f l o w - l e v e l cause and
effect, or t o e s t a b l i s h anything l i k e reasonable dose-response
curves, t h r e s h h o l d s , or to v a l i d a t e details of i n d i v i d u a l and
s y n e r g i s t i c e f f e c t s of v a r i o u s p o l l u t a n t s .
The i m p l e m e n t a t i o n of the 1958 "Delaney C l a u s e " (2) o f t h e
Food A d d i t i v e Amendments t o the F e d e r a l Food, Drug and Cosmetic
A c t was a l s o w i d e l y c o n s i d e r e d a t t h i s time as a p o s i t i v e p o l i t i c a l
precedent. The Delaney C l a u s e was interpreted f o r many y e a r s
as banning a b s o l u t e l y , from any f o o d consumed by man, any c h e m i c a l
found t o i n d u c e c a n c e r i n a n i m a l s , w i t h o u t r e f e r e n c e t o dose.
Its straightforward language made r e g u l a t o r y a c t i o n simple and
prompt, and i t was w i d e l y c r e d i t e d w i t h a d d i n g a v e r y p o s i t i v e
l e v e l of p r o t e c t i o n f o r t h e U.S. f o o d consumer. Only i n t h e l a t e
s e v e n t i e s and e i g h t i e s d i d i t b e g i n t o be a p p a r e n t t h a t d e t e c t i o n
l i m i t s were expanding so r a p i d l y t h a t c h e m i c a l t r a c e s f a r below
t h e c l e a r c a n c e r - c a u s i n g l e v e l s c o u l d be r o u t i n e l y d e t e c t e d . In

t h i s new t e c h n i c a l w o r l d , t h e r e was much l e s s u t i l i t y t o the

s t r a t e g y o f assuming a d e c r e m e n t a l human h e a l t h impact as a r e s u l t
o f any c o n n e c t i o n whatsoever o f t h e c h e m i c a l t o a n i m a l cancer.
(3)
S i m i l a r l y , t h e s t r a t e g y o f s i m p l i s t i c a l l y f o c u s i n g on human
h e a l t h t o t h e e x c l u s i o n o f o t h e r a i r p o l l u t i o n i m p a c t s , and o f
c o n s i d e r i n g i t as an a b s o l u t e w i t h no a c c e p t a b l e l e v e l o f r i s k ,
a t f i r s t seemed t o work m a r v e l o u s l y i n t h e c l e a n a i r d e c i s i o n - m a k i n g
process. The C l e a n A i r l e g i s l a t i o n was p e r e n n i a l l y p o p u l a r w i t h
the p u b l i c , and enforcement was v i g o r o u s i n t h e '70's. Major
b e n e f i t s were a l s o c l e a r : numbers o f days o f some t y p e s o f urban
p o l l u t i o n e p i s o d e s d e c l i n e d d r a m a t i c a l l y , and v i s i b l e m a n i f e s t a t i o n s
of a i r pollution effects sharply decreased. Even i n s t r i c t l y
bureaucratic arenas, t h e "human health only" approach seemed
successful. In t h e l a t e 70's, w h i l e several federal agencies
f e a r e d b e i n g swallowed
Resources, the Environmenta
out o f t h a t p o o l by d e c l a r i n
h e a l t h agency".
The problem came, o f c o u r s e , when m o n i t o r i n g and a i r q u a l i t y
measurements o f p o l l u t a n t s began to outstrip the c a p a b i l i t i e s
o f h e a l t h d a t a t o show d r a m a t i c e f f e c t s . Suddenly i t was o b v i o u s
t h a t some b a l a n c e o f r i s k s , and c o s t s and/or b e n e f i t s , would have
t o be made. Moreover, i t became a p p a r e n t t h a t t h e impacts o f
pollutants were not independent quantities, with unique
dose-response curves. I n s t e a d , many were shown t o be c l o s e l y
interdependent and c o u l d o n l y be c o n s i d e r e d i n t h e c o n t e x t o f
a l l the others. Most s i g n i f i c a n t o f a l l , as t h e l o n g range and
p e r v a s i v e damage o f a c i d r a i n and o t h e r e c o l o g i c a l e f f e c t s began
t o be p e r c e i v e d , i t was c l e a r t h a t l e g i s l a t i v e , r e g u l a t o r y , and
r e s e a r c h programs a d d r e s s e d solely t o human h e a l t h g e n e r a t e d a
s e t o f p r i o r i t i e s i n r e g u l a t o r y law t h a t c o u l d be f a r from t h o s e
r e a l l y needed f o r o v e r - a l l s o c i e t a l b e n e f i t s .
D e s p i t e t h e t r a p c r e a t e d by t h e o r i g i n a l d e c i s i o n t o t a c t i c a l l y
use a b s o l u t e human h e a l t h d a t a t o dominate i n a p o l i t i c a l argument
about a i r p o l l u t i o n c o n t r o l , t h e l o n g - s t a n d i n g a c i d r a i n debate
i s an example o f a p r o c e s s which i l l u s t r a t e s both ends o f t h e
spectrum o f use o f "facts". There have been periods when
representatives of utilities, coal companies, environmental
s c i e n t i s t s , midwestern c o a l s t a t e s , n o r t h e a s t e r n f o r e s t and s p o r t
f i s h i n g a r e a s , and r e g u l a t o r y a g e n c i e s were a b l e t o s i t t o g e t h e r
and j o i n t l y r e v i e w d a t a as w e l l as agree on f a c t u a l needs f o r
future decisions. However, t h e r e have a l s o been p e r i o d s when
these same parties have chosen t o take t h e same body of
measurements, selectively pull out " f a c t s " , package them with
d r a m a t i c or m a n i p u l a t i v e a d j e c t i v e s , and use them a l m o s t e x c l u s i v e l y
t o attempt t o stampede p u b l i c or p o l i t i c a l opinion. Industry
groups i n t h e s e p a r t s o f t h e c y c l e chose t o h i g h l i g h t "facts"
such as t h e e x i s t e n c e o f l a k e a c i d i f i c a t i o n i n t h e absence o f
f o s s i l f u e l c o n b u s t i o n i n f l u e n c e s , o r improvement o f c e r t a i n c r o p
yields i n the presence of d e p o s i t i o n of s u l f u r or nitrogen
compounds. E n v i r o n m e n t a l i n t e r e s t s , i n c o n t r a s t , f o c u s e d on " f a c t s "
such as f o r e s t and l a k e t r a n s i t i o n s , o r c r u m b l i n g s t a t u a r y . Both
s i d e s p u b l i s h e d g l o s s y l i t e r a t u r e w i t h t i t l e s l i k e " F a c t s About

A c i d R a i n " , but p r e s e n t i n g v e r y d i f f e r e n t " f a c t s " i n very d i f f e r e n t

ways.
What i s a l s o remarkable i s t h a t d u r i n g t h e s e l a t t e r a d v e r s a r i a l
p e r i o d s of " f a c t " use, both s i d e s g e n e r a l l y f e l t abused by t h e
others' manipulations. Both bemoaned t h e d e t e r i o r a t i o n of the
h o n e s t y of the exchange, and both wrung hands about t h e g u l l i b i l i t y
of the p u b l i c o r v u l n e r a b i l i t y of t h e p o l i t i c a l system t o the
distortions or manipulations of the other side. Both sides
r e g r e t t e d a l s o t h a t t h e i r own l e v e l o f d i s c u s s i o n was "forced"
to a l e v e l of s i m p l i s t i c s l o g a n s o r s e l e c t i v e use o f "facts",
but e x p l a i n e d t h a t t h i s was the o n l y way t h a t they c o u l d c o u n t e r
t h e t a c t i c s of t h e i r opponents.
D e s p i t e t h e sense o f b e i n g " t r a p p e d " or dragged into the
f a c t m a n i p u l a t i o n a r e n a , however, t h e r e i s ample e v i d e n c e that
i t i s p o s s i b l e t o break out o f t h a t entrapment. B e f o r e passage
was f i n a l l y a c h i e v e d , th
languished i n several
groups on one s i d e , an group ,
t r a d i n g " f a c t s h e e t s " as they j o c k e y e d f o r p u b l i c and political
support. Not s u r p r i s i n g l y , when c o n f r o n t e d by two s e t s o f e x p e r t s
c i t i n g " f a c t s " as p r o o f of o p p o s i t e p o s i t i o n s , the p o l i t i c a l b o d i e s
t y p i c a l l y t e n d t o a v o i d d e c i s i v e a c t i o n or even any a c t i o n a t
all. TOSCA i s not t h e o n l y example o f a major p i e c e o f l e g i s l a t i o n
which was pushed enthusiastically by proponents, and bitterly
opposed by a d v e r s a r i e s , but g i v e n d e t e r m i n e d n e g l e c t by a much
l a r g e r group r a t h e r than v o t e d up o r down. In such s i t u a t i o n s ,
e n d l e s s p r o c e d u r a l s t e p s seem e n e v i t a b l y t o d r a g out because t h e r e
i s no c o l l e c t i v e w i l l t o s h o r t e n them, and somehow d e a d l i n e s o f
b r i n g i n g d e c i s i v e v o t e s a r e never met.
However, i n TOSCA's c a s e , both sides f i n a l l y acknowledged
t h e r i s k of u s i n g measurements and f a c t s p r i m a r i l y as m a n i p u l a t i v e
t o o l s , and e v e n t u a l l y s w i t c h e d t o use them as a r a t i o n a l b a s i s
for action. The r i s k of s t a l e m a t e f o r TOSCA opponents was that
t h e i r s u c c e s s i n d e l a y of any a c t i o n c o u l d have l e d t o a sudden
acute political frustration, perhaps triggered by an e x t e r n a l
e v e n t , which might l e a d t o enactment o f an extreme measure much
s t r o n g e r than t h a t which was a c h i e v a b l e by working c o n s t r u c t i v e l y
with f a c t s to b u i l d a mutually t o l e r a b l e s o l u t i o n . The r i s k f o r
proponents was j u s t the c o n v e r s e : t h e y might, i n d e e d , by e n t e r i n g
i n t o compromise p r o b l e m - s o l v i n g w i t h t h e i r opponents, have m i s s e d
a chance f o r a c h i e v i n g something c l o s e r t o t h e i r i d e a l g o a l s .
N a t u r a l l y , I c a n ' t know i n d e t a i l how t h e s e r i s k s were b a l a n c e d
i n t h e minds of a l l p a r t i c i p a n t s i n t h e p o l i t i c a l s t r u g g l e o v e r
TOSCA, acid rain, or other issues. Whatever these internal
c o n s i d e r a t i o n s , however, the r e s u l t s u s u a l l y show a d e f i n i t e p a t t e r n
o f s w i n g i n g from one approach t o f a c t s and measurements t o t h e
other. A not u n f a m i l i a r p a t t e r n i n t h e p o l i t i c a l system i s t h a t
of initial fact and measurement m a n i p u l a t i o n and exploitation
t o s t a k e out p o s i t i o n , and l a t e r i n t e l l i g e n t use t o b u i l d r e a s o n a b l e
p o s i t i o n s a f t e r the f a t i g u e and f u t i l i t y of a d v e r s a r y " f a c t " games
have s e t i n . A common v a r i a t i o n i s t h a t o f i n i t i a l c o o p e r a t i v e
work w i t h a common base of f a c t and measurement among t e c h n i c a l
colleagues, which switches to the adversary position when
professional " a d v e r s a r y coaches" ( l o b b y i n g and p u b l i c relations

firms, certain kinds of general counsel offices, etc) become

involved. And though t h i s p a t t e r n may eventually evolve, due
to f a t i g u e or other reasons, back t o t h e c o o p e r a t i v e f a c t - u s e
mode, t h e a d v e r s a r y mode i n many d e b a t e s ( l i k e the c o n t i n u i n g
i n a b i l i t y t o r e a c h a up-dated v e r s i o n o f t h e C l e a n A i r A c t ) , may
s i m p l y c o n t i n u e i n d e f i n i t e l y w i t h a few minor accomodations o r
a c t i o n s , but c o n t i n u a l u n d e r l y i n g b i t t e r " f a c t " m a n i p u l a t i o n back
and f o r t h .
Another p a r t i c u l a r l y c l a s s i c example o f t h i s c y c l e u n f o l d e d
concurrently with t h i s w r i t i n g . A f t e r a fourteen year stalemate
of p o s t u r i n g and a d v e r s a r y c o n f r o n t a t i o n s , Congress appeared to
be ready to substantially revise federal pesticide regulatory
law (the Federal Insecticide, Fungicide, and Rodenticide Act
(FIFRA)) w i t h t h e s u p p o r t o f both major i n d u s t r y and environmental
groups. " F a c t s " and "measurements" had been, f o r seven C o n g r e s s e s ,
merely pawns i n a c o n t i n u i n g s t r u g g l e i n which each s i d e was
c o n v i n c e d o f the o t h e r s
of t h e i r narrow g r e e d an
had been c i t e d e n d l e s s l y over t h e c o u r s e o f t h e f o u r t e e n y e a r s ,
and yet these "facts" and measurements a t l e a s t superficially
had p r o v i d e d l i t t l e h e l p i n f o r m i n g n a t i o n a l p o l i c y . They were
d e s i g n e d o r used t o upstage o r negate, r a t h e r t h a n b u i l d on each
other. However, by O c t o b e r o f 1986 b o t h t h e House and Senate
had v o t e d overwhelmingly (329-4 i n t h e House) t o b r i n g f o r t h a
compromise v e r s i o n o f t h e l e g i s l a t i o n .
What caused t h e change i n t h e s t y l e o f t h e FIFRA d e b a t e ? As
u s u a l i n most major c o n t r o v e r s i e s , t h e dynamics o f t h e p e r i o d s
of t e n s i o n , o r o f p o s i t i v e a t t i t u d e s , a r e h a r d t o t r a c e . In t h i s
c a s e , as i s o f t e n t r u e , t h e r e was an e x t e r n a l f a c t o r which drove
t h e p a r t i e s toward compromise, and i n so d o i n g m o t i v a t e d them
to use their data and measurements c o n s t r u c t i v e l y r a t h e r than
i n a gaming p a t t e r n . I n d u s t r y groups i n t h i s c a s e were v e r y much
interested in legislative e x t e n s i o n of t h e i r patent r i g h t s to
compensate f o r l o n g r e g u l a t o r y d e l a y s ; e n v i r o n m e n t a l group s u p p o r t
was needed f o r t h i s and thus t h e r e was a potential for barter
between t h e elements o f p e s t i c i d e r e g u l a t o r y and p a t e n t l e g i s l a t i o n .
In a b r o a d e r sense, however, as t h e New York Times r e p o r t e d ( 4 ) :
"The changes in the law reflect growing scientific
knowledge and public awareness about the dangers of
pesticides. In r e c e n t y e a r s i n c r e a s i n g l y advanced t e s t i n g
has found t h a t some p e s t i c i d e s can cause c a n c e r , b i r t h
d e f e c t s and m u t a t i o n s i n humans."
The important o b s e r v a t i o n i s p r o b a b l y t o note t h a t d u r i n g
t h e f o u r t e e n y e a r s o f p i t t i n g " f a c t " a g a i n s t " f a c t " , measurement
a g a i n s t measurement, t h e two s i d e s d i d not e x a c t l y balance to
z e r o knowledge g a i n e d . However f r u s t r a t i n g i n appearance t o t h e
scientific community or public on both sides, retrospectively
one can see t h a t slow p r o g r e s s was b e i n g made i n e s t a b l i s h i n g
a base o f knowledge and a body of " a c c e p t e d f a c t " as opposed t o
t h e i s o l a t e d a n e c d o t a l " f a c t s " used o n l y f o r d e b a t i n g p u r p o s e s .
The FIFRA revisions eventually failed in the rush for
Congressional adjournment, because of House-Senate Conference
disagreement on technicalities of the patent p r o v i s i o n s . Only
t h e coming Congress w i l l t e l l us whether a s t a b l e consensus has

been a c h i e v e d , which can be r e s u r r e c t e d and l e g i s l a t e d again,

o r whether a n o t h e r c y c l e o f " f a c t " m a n i p u l a t i o n w i l l be n e c e s s a r y
to the process.
The Risks of Misunderstanding the Debate Cycles;
I've anecdotally documented t h e c y c l e o f measurement and f a c t

use i n p u b l i c p o l i c y d e b a t e , b u t need t o add a f u r t h e r o b s e r v a t i o n
on t h e need f o r awareness o f t h i s c y c l e , and r i s k s o f n a i v e t e
c o n c e r n i n g i t . I n an i d e a l w o r l d we might hope t h a t a l l c o n c e r n e d
would be w i l l i n g t o plunge i m m e d i a t e l y i n t o a c o n s t r u c t i v e and
purely o b j e c t i v e mode o f u s i n g a body o f measurement d a t a t o
optimize policy courses. Various segments of the s c i e n t i f i c
community p e r i o d i c a l l y e x p r e s s t h e hope f o r t h i s ideal world,
with f r u s t r a t i o n a t the true state of a f f a i r s . There c e r t a i n l y
i s no f a u l t i n b e i n g d i s a p p o i n t e
s t a t e of p u b l i c debate
it i s crucial t o understan y posturing
a n a t u r a l and even i n e v i t a b l e human group (and t h u s p o l i t i c a l )
b e h a v i o r , and n o t an i n d i c a t i o n t h a t a more r a t i o n a l o r c o n s t r u c t i v e
approach i s h o p e l e s s . One r i s k o f n a i v e t e i s to find oneself
i n a s t a t e o f shock i n r e a c t i o n t o t h e a d v e r s a r y phase, l e a d i n g
t o d r o p - o u t from t h e p r o c e s s . T h i s can r e s u l t i n n o n - p a r t i c i p a t i o n
a t t h e c r i t i c a l time i n t h e c y c l e when p a r t i c i p a n t s a r e r e a d y ,
for whatever reason, to swing back to the constructive
p r o b l e m - s o l v i n g mode. The c o n v e r s e r i s k i s t o f i n d o n e s e l f s p e n d i n g
an inordinate amount o f time speaking i n objective scientific
language i n a d i s c u s s i o n which i s p r o c e e d i n g w i t h an e n t i r e l y
d i f f e r e n t approach. For the t e c h n i c a l person not tuned t o the
a d v e r s a r y s t y l e , i t may be b e s t t o s i m p l y w a i t o u t t h e a d v e r s a r y
p e r i o d s , l i k e t i m e s o f stormy weather, b u t i n a s t a t e o f r e a d i n e s s
f o r t h e moment when t h e storm blows o v e r .
On t h e o t h e r s i d e a r e t h o s e so committed t o a f u n d a m e n t a l l y
a d v e r s a r y approach t o p o l i t i c s t h a t t h e y may c o m p l e t e l y f a i l t o
recognize an o p p o r t u n i t y f o r t r u e win-win compromise w i t h their
opponents. Though t h e hardened a d v e r s a r y a t t i t u d e may seem o f t e n
t o be " r e a l i s t i c " , i t s p r o p o n e n t s r i s k b e i n g t o t a l l y d i s e n f r a n c h i s e d
from p a r t i c i p a t i o n i n eventual compromise when t h e c o u r s e o f
decision-making turns i n that d i r e c t i o n . In a l l of the debates
I've mentioned; c l e a n a i r , a c i d r a i n , p e s t i c i d e r e g u l a t i o n , and
t o x i c c h e m i c a l s , I've seen b o t h i n d u s t r y and e n v i r o n m e n t a l groups
which had h e l d t o o l o n g t o r i g i d a d v e r s a r y r o l e s be s i m p l y left
out o f t h e f i n a l d i s c u s s i o n s which l e d t o compromise a c t i o n . The
l e s s o n i s simply t h a t though one must guard a g a i n s t n a i v e t e i n
expectation f o r o b j e c t i v i t y and c o n s t r u c t i v e attitudes, i t is
e q u a l l y n a i v e n o t t o remain v i g i l a n t and ready f o r o p p o r t u n i t i e s
f o r gain through cooperation and t h e h o n e s t mode o f s e r i o u s l y
i n t e g r a t i n g s c i e n t i f i c measurement and f a c t i n t o d e c i s i o n - m a k i n g .
Can Participants Control the Debate Reality i n Which They Operate?
The hardest p a r t t o argue o f my i n i t i a l t h e s i s i s t h a t p u b l i c

p o l i c y debate p a r t i c i p a n t s can t o an a p p r e c i a b l e extent "create
t h e i r own r e a l i t y " i n t h e tone o f t h e p o l i t i c a l debate. The

n e g a t i v e s i d e o f t h i s seems most p l a u s i b l e : i n early acid rain

d i s c u s s i o n s i t was p a r t i c u l a r l y o b v i o u s when many i n d u s t r y groups
began t o t o t a l up p o t e n t i a l c o n t r o l c o s t s and s w i t c h t o a s t r a t e g y
of embarrassing r a t h e r than w o r k i n g w i t h c o n t r o l p r o p o n e n t s , and
when e n v i r o n m e n t a l groups r e a l i z e d t h a t a c i d r a i n c o n c e r n s were
l i k e l y t o be a major p o l i t i c a l f o r c e i n m a i n t a i n i n g o r s t r e n g t h e n i n g
o t h e r a s p e c t s o f the C l e a n A i r A c t and t u r n e d away from an a p p r o a c h
of compromise o r i n c r e m e n t a l p i l o t a p p r o a c h e s . The a c t i o n s of
e i t h e r s i d e were s u f f i c i e n t t o c r e a t e a r e a l i t y of a b i t t e r d e b a t e
i n which measurements and " f a c t s " were weapons r a t h e r than t o o l s
for f i n d i n g optimal s o l u t i o n s .
On t h e p o s i t i v e s i d e , i t i s not easy t o argue c o n v i n c i n g l y
t h a t any p e r s o n o r groups, no m a t t e r how w e l l - m o t i v a t e d , has t h e
power t o s i n g l e - h a n d e d l y c o n v e r t an a d v e r s a r y confrontation to
a c o n s t r u c t i v e d i s c u s s i o n . What my own e x p e r i e n c e does t e l l me,
however, i s t h a t i n th
very r e a l b e n e f i t s to conservin
d i s s i p a t e d i n adversary , alway y
though u n p r e d i c t a b l e appearance of external factors which may
increase the self interest of those involved i n creating a
constructive problem-solving process.
Thus, the p a r t i c i p a n t i n such a debate who d e c i d e s t o move
from t h e a d v e r s a r y c o u r s e t o t h a t o f u s i n g measurement and facts
in a c o n s t r u c t i v e manner i s u s u a l l y not w i t h o u t allies. The
s e l f - i n t e r e s t o f t h e o t h e r s i d e , and p o w e r f u l e x t e r n a l i n f l u e n c e s
may a t any moment push i n t h e same d i r e c t i o n . C r e a t i n g the r e a l i t y
of a c o n s t r u c t i v e debate may thus t a k e a p e r s i s t e n c e and p a t i e n c e ,
but n o n e t h e l e s s t h e o p p o r t u n i t y i s v e r y l i k e l y t o come a l o n g i f
i t i s s e r i o u s l y sought. As mentioned above, the f i n a l House v o t e
this fall on the 14-year stalemated pesticide legislation was
329 t o 4. I t i s a v i v i d example o f t h e k i n d of consensus on t h e
meaning o f measurements and d a t a t h a t can be e v e n t u a l l y found
i f commitment t o do so i s s u f f i c i e n t l y s t r o n g and e n d u r i n g .
The l e s s o n o f t h e s e a c t u a l c a s e s i s t h a t we i n t h e s c i e n t i f i c
community must r e a l i z e that public policy debates have c y c l e s
of o b j e c t i v i t y and d i s t o r t i o n , i n w h i c h t h e l a t t e r i s not a s i g n
of the d i s i n t e g r a t i o n o f the p o l i t i c a l p r o c e s s but o n l y o f i t s
human-centered n a t u r e . Measurement and d a t a t h a t may seem wasted
or p r o s t i t u t e d a t a low p o i n t i n t h e d e b a t e p r o c e s s may, however,
be quietly moving a center of consensus toward a point and
c r e d i b i l i t y which w i l l l a t e r be t h e c o r e o f sound d e c i s i o n s . T h i s
i s the e n d u r i n g r e a s o n f o r t h e v i t a l need f o r measurement e x p e r t s
and other s c i e n t i s t s as p a r t i c i p a n t s i n t h e p o l i t i c a l process.
Though sometimes ignored, distorted, or abused, they are the
u l t i m a t e key t o p r o g r e s s i n r e a c h i n g sound p u b l i c p o l i c y .
Literature Cited:
1. U.S. Senate Report 91-1196, 91st Congress, 2nd Session (1970)
2. Food Additive Amendment, 1958, Federal Food, Drug and Cosmetics
Act (Sec. 409(c)(3)(A))
3. "Stretching Delaney Till It Breaks", Richard M. Cooper,
Regulation, November/December 1985, pg. 11-17,41
4. New York Times, September 20, 1986
RECEIVED September 28, 1987

Chapter 4
Estimation of Detection Limits

for Environmental Analytical
Procedures
A Tutorial
1
Cliff J. Kirchmer
Ecology & Environment, Inc., Suite 404, Cloverleaf Building 3,
A review o
the v a r i a b i l i t y of blank responses is
the preferred basis for defining
the lower l i m i t s of measurement.
The v a r i a b i l i t y of blank responses has
been used in England to estimate the
limits of detection for o f f i c i a l methods
of water analysis. In the United States,
the v a r i a b i l i t y of sample or standard
responses has been more often used to
estimate limits of detection. Practices
also differ with respect to whether or
not blank correction is done. These
practices are compared and recommenda-
tions made regarding the most appropriate
procedures for estimating lower limits
of measurement for several types of
environmental analyses.
D u r i n g t h e l a s t 20 o r so y e a r s t h e r e h a v e b e e n
s i g n i f i c a n t a d v a n c e s i n the t h e o r y f o r d e f i n i n g t h e l o w e r
l i m i t s of measurement. U n f o r t u n a t e l y , i n p r a c t i c e we
have n o t a l w a y s a p p l i e d what t h e o r y has t o l d us. T h e o r y
t e l l s us t h a t t h e v a r i a b i l i t y o f b l a n k r e s p o n s e s i s t h e
p r e f e r r e d b a s i s f o r d e t e r m i n i n g when t h e sample r e s p o n s e
i n d i c a t e s that the determinand* c o n c e n t r a t i o n i s g r e a t e r
than zero. H e i n r i c h K a i s e r was one o f t h e f i r s t s c i e n -
t i s t s t o r e c o g n i z e t h a t f a c t , and he l i k e n e d t h e s i t u -
a t i o n t o t h a t o f " s e a r c h i n g f o r a s h i p i n a stormy s e a .
Is i t a s h i p or a h i g h e r wave t h a n u s u a l ? " ( _ 3 ) I n t h i s
a n a l o g y , t h e h e i g h t o f t h e waves i n c o m p a r i s o n w i t h t h e
*"Determinand" means that which i s to be determined (1,2)-

1
Current address: Manchester Laboratory, State of Washington Department of Ecology, P.O.
Box 346, Manchester, WA 98353
0097-6156/88/0361-0078$06.00/0

4. KIRCHMER Estimation for Environmental Analytical Procedures 79
h e i g h t o f t h e s h i p l i m i t t h e a b i l i t y o f one t o d e t e c t t h e
p r e s e n c e o f t h e s h i p , j u s t as t h e v a r i a b i l i t y i n t h e
b a c k g r o u n d ( i . e . b l a n k ) r e s p o n s e l i m i t s the a b i l i t y to
d e t e c t the p r e s e n c e of the d e t e r m i n a n d . Note t h a t i t i s
not t h e d e p t h o f t h e o c e a n t h a t l i m i t s t h e a b i l i t y t o
d e t e c t the p r e s e n c e of the s h i p , j u s t as i t i s not the
a b s o l u t e b l a n k r e s p o n s e t h a t l i m i t s the a b i l i t y to d e t e c t
the p r e s e n c e of d e t e r m i n a n d . A c c o r d i n g to K a i s e r , "the
c a u s e of the u n c e r t a i n t y i n the a n a l y t i c a l v a l u e i s not
due t o t h e s i z e i t s e l f o f t h e b l a n k m e a s u r e , b u t t o t h e
s i z e of the f l u c t u a t i o n s i n i t . A c o n s t a n t b l a n k measure
of w h a t e v e r s i z e can a l w a y s be compensated."(3) Thus, i n
d e t e r m i n i n g the c r i t e r i o n of d e t e c t i o n one must measure
the v a r i a b i l i t y of b l a n k r e s p o n s e s , and i n d e t e r m i n i n g
whether a sample r e s p o n s e i n d i c a t e s the p r e s e n c e of
d e t e r m i n a n d one must f i r s t t fo blank
Then, the b l a n k c o r r e c t e
the c a l c u l a t e d c r i t e r i o
the d e t e r m i n a n d has been d e t e c t e d . The e x p r e s s i o n c r i -
t e r i o n o f d e t e c t i o n has b e e n u s e d h e r e i n a g e n e r a l
s e n s e , but w i l l be more p r e c i s e l y d e f i n e d i n the n e x t
section.
Variability of the B l a n k Responses
S e v e r a l a u t h o r s have p u b l i s h e d p a p e r s r e g a r d i n g the
c a l c u l a t i o n of the l i m i t of d e t e c t i o n based on the v a r i -
a b i l i t y of the b l a n k r e s p o n s e s . P r o m i n e n t among t h e s e
have been H. Kaiser(_3) i n Germany, A.L. W i l s o n ^ ) i n
E n g l a n d , and L. Currie(_5) i n the U.S.A. The f o l l o w i n g
t h e o r e t i c a l t r e a t m e n t i s based on the work of A.L.
Wilson. The c o n c l u s i o n s of W i l s o n a r e s i m i l a r to t h o s e
of K a i s e r and C u r r i e , the p r i m a r y d i f f e r e n c e s b e i n g the
t e r m i n o l o g y and the c h o i c e s made f o r e r r o r s of the 1 s t
and 2nd k i n d s .
Cheeseman and W i l s o n have s t a t e d t h a t " i t i s a w e l l -
e s t a b l i s h e d c o n c e p t t h a t each method has a l o w e r c o n c e n -
t r a t i o n l i m i t b e l o w which the d e t e r m i n a n d c a n n o t be de-
tected. However, t h e r e i s a g r e a t d e a l of v a r i a b i l i t y i n
how t h i s l i m i t i s c h o s e n and i n how r e s u l t s a r e r e p o r t e d
when the d e t e r m i n a n d was not detected."(6^) W i l s o n p r o -
posed d e t e r m i n i n g the l i m i t of d e t e c t i o n based on the
v a r i a b i l i t y of the b l a n k as a means of c o n t r i b u t i n g
u n i f o r m i t y as w e l l as a c c u r a c y t o the r e p o r t i n g of r e -
s u l t s a t low c o n c e n t r a t i o n s . W i l s o n , f o l l o w i n g K a i s e r on
t h i s m a t t e r , emphasized t h a t " t h e b l a n k s h o u l d g e n e r a l l y
be a n a l y z e d by e x a c t l y t h e same p r o c e d u r e as t h a t u s e d
for samples. T h i s s i m p l e and o b v i o u s c o n c l u s i o n i s
worth s t a t i n g because i t a p p e a r s o f t e n to be i g n o r e d . Of
c o u r s e , s i t u a t i o n s a r i s e where i t i s i m p r a c t i c a l or not
e s s e n t i a l to a n a l y z e b l a n k s and s a m p l e s i d e n t i c a l l y , but
such s i t u a t i o n s g e n e r a l l y r e q u i r e e x p e r i m e n t a l c o n f i r m -
a t i o n . "(6)

C r i t e r i o n of D e t e c t i o n . A c c o r d i n g t o W i l s o n , an a n a l y t i -
c a l r e s u l t , R, i s e q u a l t o t h e s a m p l e r e s p o n s e m i n u s t h e
b l a n k r e s p o n s e ( i . e . R = S - B). T h i s i s t h e ' p a i r e d -
c o m p a r i s o n ' s i t u a t i o n , i n which we compare i n d i v i d u a l
sample and b l a n k r e s p o n s e s . I f we were to d e t e r m i n e t h e
d i s t r i b u t i o n o f r e s u l t s when sample and b l a n k a r e i d e n t i -
c a l ( i . e . sample does n o t c o n t a i n t h e d e t e r m i n a n d ) , we
w o u l d o b t a i n t h e d i s t r i b u t i o n r e p r e s e n t e d i n F i g u r e 1.
In t h i s d i s t r i b u t i o n , t h e mean r e s u l t i s z e r o w i t h e q u a l
numbers o f p o s i t i v e and n e g a t i v e r e s u l t s d i s t r i b u t e d
around the mean. A n o r m a l d i s t r i b u t i o n i s assumed. In
o r d e r to c o n c l u d e t h a t the sample c o n t a i n s the d e t e r m i -
n a n d we must c h o o s e t h e l e v e l o f r i s k we a r e p r e p a r e d t o
take i n c o m m i t t i n g an e r r o r of t h e 1 s t k i n d ( t h a t i s , a
f a l s e p o s i t i v e , c o n c l u d i n g t h a t the d e t e r m i n a n d has been
d e t e c t e d when i n f a c t none i s p r e s e n t ) W i l s o n chose a
l e v e l of 5% ( a=0.05
F i g u r e 1, t h i s c h o i c
c r i t e r i o n o f d e t e c t i o n e q u a l t o 1.645( / 2 ) a R or 2.33a B
(where i s the w i t h i n - b a t c h s t a n d a r d d e v i a t i o n o f t h e
b l a n k r e s p o n s e ) , meaning t h a t v a l u e s g r e a t e r t h a n 2.33
a r e c o n s i d e r e d to i n d i c a t e t h a t t h e d e t e r m i n a n d has been
d e t e c t e d , w i t h 1 chance i n 20 o f b e i n g wrong.
L i m i t of D e t e c t i o n . One must a l s o c o n s i d e r t h e p o s s i b i l i -
t y o f e r r o r s of the s e c o n d k i n d ( t h a t i s , f a l s e n e g a t i v e s
or t h e p r o b a b i l i t y o f f a l s e l y c o n c l u d i n g t h a t t h e sample
does not c o n t a i n d e t e r m i n a n d , when i n f a c t i t i s p r e -
sent). F o r a sample whose t r u e c o n c e n t r a t i o n i s e q u a l t o
the c r i t e r i o n of d e t e c t i o n , t h a t p r o b a b i l i t y i s e q u a l t o
50%. W i l s o n c h o s e t o r e d u c e t h a t v a l u e t o 5%, as i l l u s -
t r a t e d i n F i g u r e 2. The l i m i t o f d e t e c t i o n i s d e f i n e d as
b e i n g t w i c e t h a t o f the c r i t e r i o n o f d e t e c t i o n , o r
4.65ag . Thus, t h e l i m i t of d e t e c t i o n i s t h e s m a l l e s t
sample c o n c e n t r a t i o n t h a t can be d e t e c t e d w i t h 95%
probability.
Discussion. The r e l a t i o n s h i p o f t h e c r i t e r i o n of d e t e c -
t i o n to the l i m i t of d e t e c t i o n i s i l l u s t r a t e d i n F i g u r e
3. In a p p l y i n g t h e s e c o n c e p t s t o sample r e s u l t s , t h e
d e t e r m i n a n d i s c o n s i d e r e d d e t e c t e d i f t h e sample r e s u l t
i s e q u a l t o or g r e a t e r t h a n t h e c r i t e r i o n o f d e t e c t i o n .
However, i f a r e s u l t i s l e s s than the c r i t e r i o n of d e t e c -
t i o n , i t i s r e p o r t e d as l e s s t h a n t h e l i m i t o f d e t e c t i o n
to take i n t o a c c o u n t t h e p o s s i b i l i t y o f f a l s e n e g a t i v e s .
Suppose, f o r example, t h a t t h e c r i t e r i o n o f d e t e c t i o n ,
C , i s e q u a l t o 5 and t h e l i m i t of d e t e c t i o n , L , i s
n n
e q u a l t o 10. Then a r e s u l t o f 3 would be r e p o r t e d as<10,

w h i l e a r e s u l t o f 7 w o u l d be r e p o r t e d as s u c h ( i . e . 7 ) .
W h i l e c o n s i s t e n t w i t h t h e d e f i n i t i o n s , t h e r e s u l t s have
r a r e l y been r e p o r t e d i n t h i s manner. In o r d e r t o a v o i d
the l o s s of i n f o r m a t i o n , two a l t e r n a t i v e ways of r e p o r t -
i n g r e s u l t s a t s m a l l c o n c e n t r a t i o n s have been s u g g e s t e d .
C u r r i e ( 7 ^ has s u g g e s t e d t h a t r e s u l t s b e l o w t h e c r i t e r i a
of d e t e c t i o n be r e p o r t e d as N.D. ( n o t d e t e c t e d ) , but t h a t

Mean, u = 0
Criterion of Detection = 2.33aB

3 j (Shaded Area Denotes 5%
of Ail Results)
ca
<D
DC
-2a -a 0 o'
Difference Between Two Blanks
Figure 1 D i s t r i b u t i o n of the D i f f e r e n c e s Between

A n a l y t i c a l R e s p o n s e s f o r Two B l a n k s -
D e f i n i t i o n of the C r i t e r i o n of D e t e c t i o n
( A d a p t e d w i t h p e r m i s s i o n from Ref. 6.
C o p y r i g h t 1978 Water R e s e a r c h C e n t r e ) .

L -3a'
D Lo-2a' l^-a' L,, Lâ Lâ' Lâ'
Difference Between a Sample and a Blank
Figure 2 D i s t r i b u t i o n of t h e D i f f e r e n c e s Between
A n a l y t i c a l R e s p o n s e s f o r a Sample and a
B l a n k When t h e Sample C o n c e n t r a t i o n i s
E q u a l t o The D e f i n e d L i m i t o f D e t e c t i o n
(Adapted w i t h p e r m i s s i o n from Ref. 6.
C o p y r i g h t 1978 Water R e s e a r c h C e n t r e ) .

?
S.
.
*
I
s.
"
00
w
t
Co
Co
3
?

the e s t i m a t e d v a l u e a l s o be r e p o r t e d . ( i . e . i n the
example a b o v e , the e s t i m a t e d v a l u e of 3 w o u l d be r e p o r t -
ed, a l o n g w i t h a not d e t e c t e d d e c i s i o n ) . Hunt and
W i l s o n ( 8 ^ have s u g g e s t e d t h a t a l l r e s u l t s at low c o n c e n -
t r a t i o n s s i m p l y be r e p o r t e d as the r e s u l t p l u s or minus
the c o n f i d e n c e l i m i t s at a s t a t e d l e v e l of c o n f i d e n c e
(R + t s ) . T h i s r a d i c a l s u g g e s t i o n w o u l d c o m p l e t l y e l i m i -
n a t e t h e u s e o f t h e c r i t e r i o n o f d e t e c t i o n and t h e l i m i t
of d e t e c t i o n i n r e p o r t i n g r e s u l t s . However, t h e y c o u l d
be p r o v i d e d s e p a r a t e l y as r e f e r e n c e i n f o r m a t i o n f o r the
u s e r to a i d i n the i n t e r p r e t a t i o n of the r e p o r t e d r e s u l t s
and a s s o c i a t e d confidence l i m i t s .
W i l s o n used " p a i r e d o b s e r v a t i o n s " i n d e f i n i n g C n
and L . n That i s , each sample r e s p o n s e i s c o n s i d e r e d to

be i n d i v i d u a l l y blank corrected. It i s a l s o p o s s i b l e
to d e f i n e a d e t e c t i o n l i m i t based on a " w e l l - k n o w n "
blank. In t h i s c a s e
between sample and b l a n
of y/l. (5) Hunt and W i l s o n have p r e s e n t e d a g e n e r a l
expression for C based on
n n r e p l i c a t e blank responses
and m r e p l i c a t e sample r e s p o n s e s , i n which the s t a n d a r d
d e v i a t i o n of the d i f f e r e n c e between sample and b l a n k
r e s p o n s e s (a g_ ) i s g i v e n by the f o l l o w i n g e q u a t i o n :
B
a = a n m n m 1 2
S-B Bf( + )/ l ^
O n l y the more common s i t u a t i o n s of a " w e l l - k n o w n " b l a n k

or " p a i r e d - c o m p a r i s o n s " w i l l be f u r t h e r c o n s i d e r e d h e r e .
In summary, i n d e f i n i n g the d e t e c t i o n l i m i t based
on the v a r i a b i l i t y of b l a n k r e s p o n s e s , one must make two
choices. F i r s t , one must choose whether b l a n k c o r r e c -
t i o n i s t o be b a s e d on a " w e l l - k n o w n " b l a n k o r on
"paired-comparisons". Second, one must choose the
v a l u e s f o r a and 3, c o r r e s p o n d i n g to the r i s k s of e r r o r s
of the f i r s t and second k i n d s . T a b l e I summarizes some
of the d e f i n i t i o n s t h a t h a v e been p r o p o s e d by Currie.(_5)
Table I . "Working" E x p r e s s i o n s For C r i t i c a l

L e v e l , D e t e c t i o n L i m i t , and Determi-
n a t i o n L i m i t P r o p o s e d by L l o y d C u r r i e
Critical Detection Determination

Level Limit Limit
Paired
Observations 2.33 a 4.65 a 14.1 a
B B B
"Well-known
Blank 1.64 a,B 3.2 9 a,B 10 a,B

Wilson's d e f i n i t i o n s g e n e r a l l y agree with those of

C u r r i e e x c e p t t h a t he u s e s t h e terms c r i t e r i o n o f d e t e c -
t i o n and l i m i t o f d e t e c t i o n r a t h e r t h a n c r i t i c a l l e v e l
and d e t e c t i o n l i m i t , and r e s t r i c t s h i m s e l f t o a c o n s i d -
eration of p a i r e d o b s e r v a t i o n s . Currie also introduces
a new term, t h e d e t e r m i n a t i o n l i m i t , L Q , which i s
d e f i n e d as t h a t c o n c e n t r a t i o n f o r which t h e r e l a t i v e
s t a n d a r d d e v i a t i o n i s 10%. R e s u l t s above t h e d e t e r m i n -
a t i o n l i m i t may t h e r e f o r e be c o n s i d e r e d s a t i s f a c t o r y f o r
quantitative analysis. T a b l e I I compares t h e d e f i n i -
t i o n s by C u r r i e w i t h t h o s e of K a i s e r f o r " p a i r e d o b s e r -
vations".
Table II. A Comparison o f D e f i n i t i o n s by L.

C u r r i e an
Comparisons
Kaiser Currie
D e f i n i t i o n s Based L i m i t of D e t e c t i o n Critical Level

On E r r o r of t h e (Nachweisgrenze) = 2 .33G B
First Kind =4.24 ag a =0.05

a =0.0014
D e f i n i t i o n s Based L i m i t of G u a r a n t e e Detection Limit

On E r r o r s o f t h e (Garantiegrenze) = 4.65 a B
F i r s t and Second = 8.49a B

a= 3 =0.05
Kinds a = 3 =0.0014
Note t h a t t h e d i f f e r e n c e s l i e i n t h e c h o i c e o f a and &

values. The i m p o r t a n t t h i n g i s n o t how t h e s e d e f i n i -
t i o n s d i f f e r , but r a t h e r t h a t w h i c h t h e y h a v e i n common.
A l l a r e based on a c o n s i d e r a t i o n o f t h e v a r i a b i l i t y o f
the b l a n k and an a s s u m p t i o n t h a t sample v a r i a b i l i t y i s
the same as b l a n k v a r i a b i l i t y a t low c o n c e n t r a t i o n s .
A l l r e q u i r e some s o r t of b l a n k c o r r e c t i o n , based on
" p a i r e d o b s e r v a t i o n s " or a " w e l l - k n o w n " b l a n k . A l l are
based on d e t e r m i n i n g the l i m i t of d e t e c t i o n of a
"complete a n a l y t i c a l procedure". And f i n a l l y , a l l a r e
b a s e d on some c h o i c e o f a c c e p t a b l e e r r o r s o f t h e 1 s t
and 2nd k i n d s .
As F i g u r e 1 shows, when t h e sample does n o t c o n t a i n
the d e t e r m i n a n d , t h e r e s u l t i s n e g a t i v e h a l f o f t h e
time. But t h e s e n e g a t i v e r e s u l t s a r e n o t r e p o r t e d s i n c e
b e l o w the C a l l r e s u l t s a r e i n d i s t i n g u i s h a b l e from t h e
n
b l a n k due t o random e r r o r . However, an a n a l y t i c a l r e s u l t

i s not equal to the sample response, but r a t h e r to the
d i f f e r e n c e between sample r e s p o n s e and b l a n k r e s p o n s e .
T h i s r e s u l t , w h i c h may be p o s i t i v e , n e g a t i v e , or z e r o i s
compared w i t h C and t h e a p p r o p r i a t e d e c i s i o n r e g a r d i n g
n
d e t e c t i o n made.

From t h e d i s c u s s i o n t o t h i s p o i n t , i t would a p p e a r
t h a t we o n l y n e e d a p p l y t h e o r y t o p r a c t i c e i n o r d e r t o
a r r i v e at our d e s i r e d c r i t e r i o n or l i m i t of d e t e c t i o n .
U n f o r t u n a t e l y , the t h e o r y i s based on t h e f o l l o w i n g
a s s u m p t i o n s , w h i c h may n o t a l w a y s h o l d :
1) That t h e w i t h i n - b a t c h s t a n d a r d d e v i a t i o n s o f b o t h
the b l a n k and s a m p l e s c o n t a i n i n g v e r y s m a l l c o n c e n t r a -
t i o n s of t h e d e t e r m i n a n d a r e t h e same
2) That t h e a n a l y t i c a l r e s p o n s e i s n o t z e r o f o r
f i n i t e c o n c e n t r a t i o n s o f the d e t e r m i n a n d
3) That t h e s a m p l e and b l a n k a r e n o t b i a s e d w i t h
r e s p e c t t o each o t h e r ( t h a t i s , t h e r e a r e no i n t e r f e r i n g
s u b s t a n c e s i n the sample o r t h e b l a n k )
I f any one o f t h e a b o v e a s s u m p t i o n s i s n o t t r u e , t h e n
the l i m i t o f d e t e c t i o n c a n n o t be c a l c u l a t e d u s i n g t h e
equations given p r e v i o u s l y I however
a d j u s t m e n t s can be
example, has p r e s e n t e
c o r r e c t i o n s when a s s u m p t i o n s 1) and 3) a b o v e a r e n o t
met.(9^) F o r example, as i l l u s t r a t e d i n F i g u r e 4,
a d j u s t m e n t s c a n be made t o a l l o w f o r d i f f e r e n c e s i n t h e
s t a n d a r d d e v i a t i o n f o r b l a n k and sample r e s p o n s e s
(On f a~ ) and f o r d i f f e r e n t v a l u e s f o r e r r o r s o f t h e
1st and 2nd k i n d s . A l s o , when s y s t e m a t i c e r r o r c a n n o t
be assumed n e g l i g i b l e , t h e l i m i t o f d e t e c t i o n must be
i n c r e a s e d by an a m o u n t , 2 where i s t h e assumed
upper bound f o r the b i a s
i.e. C n + and L ' = 2 C '

D n
Existing P r a c t i c e i n Water A n a l y s i s
A f t e r t h i s b r i e f r e v i e w o f t h e o r y , l e t us t u r n o u r
a t t e n t i o n t o e x i s t i n g p r a c t i c e , as e x e m p l i f i e d i n e n v i -
r o n m e n t a l methods o f a n a l y s i s . E n v i r o n m e n t a l methods o f
a n a l y s i s employ many o f the common a n a l y t i c a l i n s t r u -
ments i n a n a l y z i n g a wide s p e c t r u m o f c h e m i c a l s i n a
v a r i e t y of m a t r i c e s . I n s t r u m e n t s commonly used i n c l u d e
spectrophotometers (atomic a b s o r p t i o n , v i s i b l e , i n d u c t -
i v e l y c o u p l e d p l a s m a ) , gas c h r o m a t o g r a p h s ( w i t h a v a r i e -
t y o f d e t e c t o r s , i n c l u d i n g t h e mass s p e c t r o m e t e r ) , and
automatic a n a l y z e r s .
V a r i a b i l i t y of Blank Responses. In o r d e r t o l i m i t t h e
d i s c u s s i o n , l e t us f o c u s on w a t e r a n a l y s e s as represen-
t a t i v e of e n v i r o n m e n t a l a n a l y s e s . In t h e U n i t e d Kingdom,
the S t a n d i n g Committee o f A n a l y s t s o f t h e Department o f
the E n v i r o n m e n t i s s u e s a n a l y t i c a l methods i n a s e r i e s o f
booklets. I n c l u d e d among t h e s e a r e the 'Methods f o r t h e
Examination o f Waters and A s s o c i a t e d Materials.'(10-14)
S e v e r a l of t h e s e methods have been e v a l u a t e d by i n d i v i d -
u a l l a b o r a t o r i e s to determine the l i m i t of d e t e c t i o n
based on t h e v a r i a b i l i t y of t h e b l a n k and u s i n g ' p a i r e d
comparisons' f o r blank c o r r e c t i o n . Published values for
the l i m i t o f d e t e c t i o n f o r s e v e r a l o f t h e s e methods a r e

l i s t e d i n T a b l e I I I . The S t a n d i n g Committee of A n a l y s t s
has a d o p t e d a p o l i c y of i n c l u d i n g an e s t i m a t e of the
l i m i t of d e t e c t i o n ( o r the w i t h i n - b a t c h s t a n d a r d d e v i -
a t i o n of the b l a n k , which i s used to c a l c u l a t e the l i m i t
of d e t e c t i o n ) as one of the 'Performance C h a r a c t e r i s t i c s
of t h e M e t h o d ' .
Table I I I . E s t i m a t e d L i m i t s of D e t e c t i o n
f o r S e l e c t e d Methods t a k e n from
"Methods f o r the E x a m i n a t i o n of Waters
and A s s o c i a t e d M a t e r i a l s "
Estimated Limit Degrees of Freedom

Element of D e t e c t i o n * * Used i n E s t i m a t e ( s )
Copper ( 1 0 ) * 1 .7-
, . ug/L
Chromium ( 1 1 ) * 3,.2-7/4 ug/L 7-9
Phosphorus ( 1 2 ) * 0,.003 -0 .006 mg/L 35
S i l i c o n (13)* 0,.03 mg/L 10
Aluminum ( 1 4 ) * 0,.013 m g/L 10
* N o t e - - T h e s e a r e r e f e r e n c e numbers
* * N o t e R a n g e f o r some e l e m e n t s i s due to outcomes i n
in different laboratories.
In the U n i t e d S t a t e s t h e r e a r e s e v e r a l p u b l i s h e d
methods f o r the a n a l y s i s of w a t e r s . These i n c l u d e
methods p u b l i s h e d by the E n v i r o n m e n t a l P r o t e c t i o n Agency
and the 'Standard Methods f o r the A n a l y s i s of Water and
W a s t e w a t e r ' . (15) In S t a n d a r d Methods f o r the Examin-
a t i o n of Water and Wastewater t h e r e i s s u r p r i s i n g l y
l i t t l e g u i d a n c e on how to d e t e r m i n e d e t e c t i o n l i m i t s .
For flame atomic a b s o r p t i o n spectrophotometry, the de-
t e c t i o n l i m i t i s d e f i n e d as the c o n c e n t r a t i o n t h a t p r o -
duces a b s o r p t i o n e q u i v a l e n t to t w i c e the m a g n i t u d e of
the b a c k g r o u n d f l u c t u a t i o n . No m e n t i o n i s made of the
b l a n k or b l a n k c o r r e c t i o n . T h i s d e f i n i t i o n i m p l i e s an
i n s t r u m e n t d e t e c t i o n l i m i t r a t h e r than a d e t e c t i o n l i m i t
of a 'complete a n a l y t i c a l p r o c e d u r e . ' F i n a l l y , no men-
t i o n i s made o f t h e n e e d t o d e t e r m i n e t h e v a r i a b i l i t y o f
responses.
V a r i a b i l i t y of Sample R e s p o n s e s . Under the C l e a n Water

A c t the U.S. E n v i r o n m e n t a l P r o t e c t i o n Agency has pub-
l i s h e d g u i d e l i n e s e s t a b l i s h i n g t e s t p r o c e d u r e s f o r the
a n a l y s i s of p o l l u t a n t s . A r e v i e w of t h e s e methods can
g i v e us a r e a s o n a b l e p i c t u r e of e x i s t i n g p r a c t i c e i n
d e t e r m i n i n g the l i m i t of d e t e c t i o n . In A p p e n d i x B t o
P a r t 136 of the F e d e r a l R e g i s t e r a d e f i n i t i o n and p r o c e -
dure f o r the d e t e r m i n a t i o n of the method d e t e c t i o n l i m i t
i s given.(16). The method d e t e c t i o n l i m i t (MDL) i s

d e f i n e d as "the minimum c o n c e n t r a t i o n of a s u b s t a n c e
t h a t can be i d e n t i f i e d , measured and r e p o r t e d w i t h 99%
c o n f i d e n c e t h a t the a n a l y t e c o n c e n t r a t i o n i s g r e a t e r
t h a n z e r o and i s d e t e r m i n e d f r o m a n a l y s i s o f a s a m p l e i n
a g i v e n m a t r i x c o n t a i n i n g the a n a l y t e " . The term method
d e t e c t i o n l i m i t i s a misnomer, s i n c e the v a l u e w i l l
depend on the i n s t r u m e n t s e n s i t i v i t y , the n a t u r e of the
s a m p l e s , and t h e s k i l l o f t h e a n a l y s t , as w e l l as t h e
method. The e q u a t i o n g i v e n f o r c a l c u l a t i n g the MDL i s :
MDL = ts
Where t = the Student's t v a l u e a p p r o p r i a t e f o r a

99% c o n f i d e n c e l e v e l ( o n e - s i d e d ) and a s t a n d a r d d e v i a -
t i o n e s t i m a t e w i t h n-1 d e g r e e s of f r e e d o m and s = the
s t a n d a r d d e v i a t i o n f th replicat analyse f stan
d a r d s or s a m p l e s w i t
F i g u r e 5 d e p i c t s th
o r i g i n a l l y p r o p o s e d , the p r o c e d u r e d i d not a l l o w f o r
blank correction. The l a t e s t v e r s i o n of the p r o c e d u r e
does a l l o w f o r b l a n k c o r r e c t i o n , a l t h o u g h the s t a n d a r d
d e v i a t i o n of a n a l y t e r e s p o n s e r a t h e r t h a n the s t a n d a r d
d e v i a t i o n of the b l a n k i s used to c a l c u l a t e the d e t e c -
tion limit. The d i s t r i b u t i o n shown i n F i g u r e 5 i s t h a t
of the r e s u l t ( p r e s u m a b l y a f t e r b l a n k c o r r e c t i o n i f
necessary).
T h i s d e f i n i t i o n i s m a t h e m a t i c a l l y v e r y s i m i l a r to
t h a t p r e s e n t e d e a r l i e r based on the v a r i a b i l i t y of the
d i f f e r e n c e s between sample and b l a n k when the sample
c o n c e n t r a t i o n i s e q u a l to the c r i t e r i a of d e t e c t i o n .
The 99% c o n f i d e n c e l e v e l c o r r e s p o n d s m a t h e m a t i c a l l y to
the c a s e i n w h i c h the e r r o r of the 1st k i n d i s c h o s e n to
be 1% ( a n d t h e e r r o r o f t h e 2nd k i n d i s e q u a l t o 50 % ) .
However, the MDL i s based on the v a r i a b i l i t y of a n a l y t e
r e s p o n s e r a t h e r than b l a n k r e s p o n s e . The MDL was first
d e f i n e d f o r a p p l i c a t i o n to the a n a l y s i s of t r a c e o r g a n i c
compounds and a p p a r e n t l y was based on the c o n c l u s i o n
t h a t the f i r s t a s s u m p t i o n l i s t e d a b o v e was not met.
( i . e . t h e w i t h i n - b a t c h s t a n d a r d d e v i a t i o n s of b o t h the
b l a n k and s a m p l e s c o n t a i n i n g v e r y s m a l l c o n c e n t r a t i o n s
of d e t e r m i n a n d a r e judged not to be the same). In f a c t ,
the i m p l i c a t i o n i s t h a t t h e r e o f t e n may be no s i g n i f i -
c a n t r e s p o n s e f o r the b l a n k . Hence, no need f o r b l a n k
c o r r e c t i o n and t h e o r y t h e r e f o r e e m p h a s i z e s sample r e -
sponse .
Blank C o r r e c t i o n Omitted. T h e r e a r e two common s i t u a -

t i o n s i n w a t e r a n a l y s i s where b l a n k c o r r e c t i o n i s not
made, or a p p a r e n t l y not made, e v e n though the d e t e c t i o n
of t r a c e s of d e t e r m i n a n d s i s impor t ant. (**) The first
c o r r e s p o n d s to t h o s e p r o c e d u r e s i n w h i c h b l a n k d e t e r m i -
n a t i o n s a r e made but t h e i r r e s p o n s e s a r e a d j u s t e d t o
z e r o by i n s t r u m e n t a l s u b t r a c t i o n of a c o n s t a n t from a l l
responses. T h i s p r o c e d u r e i s e q u i v a l e n t to making
s e p a r a t e measurements of b l a n k and sample r e s p o n s e s ,

F i g u r e 4. I l l u s t r a t i o n o f the case i n w h i c h the s t a n d a r d d e v i a t i o n

f o r sample and b l a n k responses d i f f e r and i n which t h e v a l u e s chosen
f o r e r r o r s o f t h e f i r s t k i n d (a) and t h e second k i n d ( ( 3 ) a l s o d i f f e r .
(Adapted w i t h p e r m i s s i o n from Ref. 9. C o p y r i g h t 1978 W i l e y . )
O MDL
Measured Analyte Concentration
F i g u r e 5. Method d e t e c t i o n l i m i t d e p i c t e d as an e r r o r d i s t r i b u t i o n .
(Adapted from Ref. 19. C o p y r i g h t 1981 American C h e m i c a l S o c i e t y . )

s i n c e the s u b t r a c t i o n o f a c o n s t a n t from a measurement

does n o t a f f e c t i t s s t a n d a r d d e v i a t i o n . The o n l y major
d i f f e r e n c e i s t h a t t h i s p r o c e d u r e c o n c e a l s any s y s t e m -
a t i c , t e m p o r a l v a r i a t i o n s o f t h e b l a n k r e s p o n s e from
b a t c h to b a t c h . A l s o , the i n i t i a l procedure to d e t e r -
m i n e t h e v a r i a b i l i t y o f t h e b l a n k may n e e d t o be a l -
t e r e d , s i n c e no i n d e p e n d e n t b l a n k r e s p o n s e v a l u e s a r e
obtained with i n s t r u m e n t a l blank c o r r e c t i o n .
The o t h e r p r o c e d u r e i n which no b l a n k c o r r e c t i o n i s
1
used i s t h a t i n w h i c h t h e d e t e r m i n a n d s presence i s
r e v e a l e d by a peak r i s i n g above a b a c k g r o u n d ; chroma-
t o g r a p h i c t e c h n i q u e s a r e a f a m i l i a r example, and a n a l -
y s e s u s i n g an A u t o A n a l y z e r have a l s o been r e p o r t e d t o
f a l l into t h i s category. In t h e s e t y p e s o f p r o c e d u r e s ,
a b l a n k - c o r r e c t i o n i s commonly made, i n e f f e c t , by mea-
s u r i n g the h e i g h t o f th determinand k abov
the b a s e - l i n e . Thi
o n l y i f thorough t e s t i n
t i c a l r e s p o n s e above t h e b a s e l i n e when t r u e b l a n k d e t e r -
m i n a t i o n s a r e made. Sometimes t h e s e n s i t i v i t y of t h e
i n s t r u m e n t i s a d j u s t e d so t h a t r e s p o n s e s a r e n o t ob-
t a i n e d u n t i l a t h r e s h o l d c o n c e n t r a t i o n of determinand i s
reached. T h i s s i t u a t i o n s h o u l d , i f p o s s i b l e , be a v o i d e d
whenever d e t e c t i o n o f v e r y s m a l l c o n c e n t r a t i o n s i s o f
p r i m a r y i n t e r e s t by a d j u s t i n g t h e s e n s i t i v i t y so t h a t
the d i s c r i m i n a t i o n has n e g l i g i b l e e f f e c t on t h e a p p a r e n t
s i z e o f random e r r o r s . As d i s c u s s e d p r e v i o u s l y , t h e
s o l u t i o n f o r c h r o m a t o g r a p h i c methods p r o p o s e d by t h e
U.S. E n v i r o n m e n t a l P r o t e c t i o n Agency i s t o d e t e r m i n e t h e
MDL based on t h e s t a n d a r d d e v i a t i o n o f low c o n c e n t r a t i o n
sample o r s t a n d a r d r e s p o n s e s . ( 1 6 ) Hunt and W i l s o n h a v e
suggested that a p r a c t i c a l l i m i t of d e t e c t i o n f o r cases
where no b l a n k c o r r e c t i o n i s n e c e s s a r y w o u l d be t h a t
c o n c e n t r a t i o n p r o d u c i n g an a n a l y t i c a l r e s p o n s e g r e a t e r
t h a n some v a l u e d, w h i c h c a n be d e t e r m i n e d f r o m t h e
a n a l y s i s o f s t andar d s.( 8^)
The EPA C o n t r a c t L a b o r a t o r y Program under S u p e r f u n d
(CERCLA) p r o v i d e s a n o t h e r example o f how b l a n k s and
d e t e c t i o n l i m i t s a r e t r ea t ed.(!]_) (AJ?.) With r e s p e c t t o
b l a n k s , t h e s t a t e m e n t s o f work s p e c i f y t h a t t h e l a b o r a -
t o r y s h o u l d not b l a n k c o r r e c t sample r e s p o n s e s . In t h e
c a s e o f o r g a n i c s a n a l y s e s , t h e EPA e v a l u a t o r a n d / o r d a t a
a u d i t o r has t h e a u t h o r i t y t o b l a n k c o r r e c t sample r e -
s p o n s e s . In p r a c t i c e t h i s i s n e v e r done. For both
o r g a n i c s and i n o r g a n i c s , t h e a b s o l u t e b l a n k l e v e l i s
p r i m a r i l y used as a c o n t r o l t o d e t e r m i n e i f s a m p l e s need
to be r e a n a l y z e d . D e t e c t i o n l i m i t s a r e based on r e p l i -
c a t e a n a l y s e s o f a s t a n d a r d a t 3-5 t i m e s t h e r e q u i r e d
d e t e c t i o n l i m i t c o n c e n t r a t i o n . The i n s t r u m e n t d e t e c t i o n
l i m i t i s c a l c u l a t e d as b e i n g e q u a l t o 3 t i m e s t h e s t a n -
d a r d d e v i a t i o n of t h e measured v a l u e . Since blank cor-
r e c t i o n i s n o t p e r m i t t e d or n o t done, sample r e s u l t s
w i l l a l l be b i a s e d h i g h by an amount e q u a l t o t h e b l a n k
response. The a b s o l u t e b l a n k v a l u e ( a c t u a l l y u s u a l l y a
m u l t i p l e o f 5 o r 10 t i m e s t h e b l a n k v a l u e ) r a t h e r t h a n

the v a r i a b i l i t y of the b l a n k d e t e r m i n e s the lower

reporting limit.
Summary and Conclusions
I t i s a p p a r e n t from the above d e s c r i p t i o n s of o f f i c i a l

m e t h o d s o f a n a l y s i s i n t h e U n i t e d S t a t e s t h a t none o f
them s p e c i f y t h a t d e t e c t i o n l i m i t s s h o u l d be d e t e r m i n e d
based on the v a r i a b i l i t y of the b l a n k f o r a 'complete
a n a l y t i c a l procedure.' T h i s c o n t r a s t s s h a r p l y w i t h the
p o l i c y of the S t a n d i n g Committee of A n a l y s t s i n t h e
U n i t e d Kingdom, which i s to p r e f e r a b l y d e t e r m i n e the
l i m i t of d e t e c t i o n based on the v a r i a b i l i t y of the
blank. Why i s t h i s ? One r e a s o n i s the d i f f e r e n c e i n
emphasis on the b l a n k i t s e l f . W h i l e i n the U.K. t h e y
have r e q u i r e d that b l a n k c o r r e c t i o b t f th
p r o c e d u r e , i n the U.S.A
b l a n k c o r r e c t i o n bu
t i o n i n some of our o f f i c i a l methods.
In the c a s e of the U.S. E n v i r o n m e n t a l P r o t e c t i o n
Agency, i t s Method D e t e c t i o n L i m i t was d e v e l o p e d i n i -
t i a l l y f o r o r g a n i c methods of a n a l y s i s u s i n g c h r o m a t o -
graphic techniques.(19) L a t e r i t s use was s u g g e s t e d f o r
o t h e r methods f o r which b l a n k c o r r e c t i o n and a d e t e c t i o n
l i m i t based on v a r i a b i l i t y of the b l a n k i s the c o r r e c t
p r o c e d u r e to f o l l o w . W i n e f o r d n e r and Long h a v e , f o r
example, r e c o g n i z e d the i m p o r t a n c e of t h e b l a n k i n s p e c -
t r o m e t r i c determinations.(20)
A n o t h e r o b s t a c l e i s the w i d e l y h e l d b e l i e f t h a t
b i a s due t o i n t e r f e r e n c e i s so l a r g e t h a t t h e r a n d o m
v a r i a b i l i t y of the b l a n k i s n e g l i g i b l e i n c o m p a r i s o n .
T h i s may be t r u e f o r some a n a l y s e s , b u t s h o u l d n o t be
assumed to h o l d i n a l l c a s e s . I f a n y t h i n g , modern
i n s t r u m e n t s a r e becoming more s e l e c t i v e and more i n t e r -
ference free. And b i a s can a l w a y s be i n c o r p o r a t e d i n t o
the e s t i m a t i o n of the d e t e c t i o n l i m i t .
F i n a l l y , t h e r e i s the a t t i t u d e t h a t e q u a t e s the
i n s t r u m e n t r e s p o n s e to the a n a l y t i c a l r e s u l t . In t h i s
c o n t e x t , b l a n k c o r r e c t i o n and the c a l c u l a t i o n of the
v a r i a b i l i t y of the b l a n k i s not u n d e r s t o o d to be a
n e c e s s a r y a d d i t i o n a l s t e p . Too much e m p h a s i s has been
placed on f o l l o w i n g c e r t a i n r u l e s , r a t h e r than doing
what i s n e c e s s a r y to o b t a i n the b e s t p o s s i b l e e s t i m a t e
of the t r u e v a l u e .
D e t e c t i o n l i m i t t h e o r y has not been a d e q u a t e l y
a p p l i e d to p r a c t i c e f o r many common o f f i c i a l l y a p p r o v e d
or s t a n d a r d methods of a n a l y s i s . To i m p r o v e t h i s s i t u -
a t i o n , the f o l l o w i n g recommendations a r e o f f e r e d :
1. In a l l c a s e s , b l a n k c o r r e c t i o n s h o u l d be done
b e f o r e c o m p a r i n g the a n a l y t i c a l r e s u l t w i t h the c r i t e -
r i o n of d e t e c t i o n to d e c i d e i f the d e t e r m i n a n d has been
detected.
2. The p r e f e r r e d p r o c e d u r e f o r d e t e r m i n i n g d e t e c -
t i o n l i m i t i s t h a t based on the v a r i a b i l i t y of b l a n k
responses f o r a complete a n a l y t i c a l procedure. This

s h o u l d a l w a y s be a t t e m p t e d f i r s t f o r a method b e f o r e
r e s o r t i n g to a l t e r n a t i v e s . I t i s u s u a l l y r e a d i l y ap-
p l i e d i n t h e c a s e o f s p e c t r o m e t r i c methods o f a n a l y s i s ,
s i n c e t h e s e methods do n o t g e n e r a l l y e x h i b i t s i g n i f i c a n t
b i a s due t o i n t e r f e r e n c e .
3. When an a n a l y s i s i n v o l v e s a peak r i s i n g above a
b a c k g r o u n d and no t r u e b l a n k r e s p o n s e i s o b t a i n e d , an
a l t e r n a t i v e p r o c e d u r e must be used. T h i s m i g h t be done
by d e t e r m i n i n g t h e MDL from t h e s t a n d a r d d e v i a t i o n o f
low l e v e l s t a n d a r d r e s p o n s e s , o r by e s t a b l i s h i n g some
c o n c e n t r a t i o n , d, t h a t must be e x c e e d e d , based on r e -
sponses obtained f o r standards. A l w a y s work a t t h e
highest p r a c t i c a l l e v e l of s e n s i t i v i t y i n order to a v o i d
the s i t u a t i o n where r e s p o n s e i s z e r o f o r f i n i t e c o n c e n -
t r a t i o n s of determinand.
4. Instrumenta
a v o i d e d , s i n c e they
sample p r e p a r a t i o n , c l e a n u p , e t c . on t h e d e t e c t i o n
limit. In t h i s r e g a r d , d e t e c t i o n l i m i t s based on s i g n a l
to n o i s e r a t i o s h o u l d a l s o be a v o i d e d . I f used, they
s h o u l d be o b t a i n e d from a 'complete a n a l y t i c a l p r o c e -
dure' and n o t j u s t t h e i n s t r u m e n t .
5. When an e s t i m a t e of s t a n d a r d d e v i a t i o n i s r e -
q u i r e d ( e i t h e r f o r b l a n k s or low c o n c e n t r a t i o n samples)
s u f f i c i e n t r e p l i c a t e s s h o u l d be done t o o b t a i n a good
e s t i m a t e of the p o p u l a t i o n v a l u e .
6. E s t i m a t e s o f s y s t e m a t i c e r r o r s , s u c h as i n t e r -
f e r e n c e , s h o u l d be i n c l u d e d i n t h e e s t i m a t e o f d e t e c t i o n
limit.
Literature Cited
1. Wilson, A. L. Talanta 1965, 12, 701.
2. International Organization for Standardization,
ISO 6107/2-1981, ISO, Geneva, 1981 (BS 6068:
Part 1: Section 1.2: 1982).
3. Kaiser, H. Two Papers on the Limit of Detection
of a Complete Analytical Procedure;Adam Hilger,
Ltd.: London, 1968.
4. Wilson, A. L. Talanta 1973, 20, 725.
5. C u r r i e , L. A. Anal. Chem. 1968, 40, 586-93.
6. Cheeseman, R. V.; Wilson, A. L. Manual On
Analytical Quality Control For the Water Industry;
Technical Report TR66: Water Research Centre:
England, 1978.
7. Currie, L. A. In X-Ray Fluorescence Analysis of
Environmental Samples; Dzubay, T. G., Ed.; Ann
Arbor Science Publishers, Inc., 1977; Chapter 25.
8. Hunt, D . T . E . ; W i l s o n , A. L . The Chemical
Analysis of Water; 2nd edition; The Royal Society
of Chemistry: London, 1986.

9. Currie, L. A. In Treatise On A n a l y t i c a l
Chemistry, Part 1, Volume 1, 2nd edition;
K o l t h o f f , I . M . a n d E l v i n g , P. E. Eds.; John
Wiley & Sons: New York, 1978; Chapter 4.
10. Standing Committee of A n a l y s t s Copper i n
Potable Waters by Atomic Absorption Spectro
photometry ,1980; Her Majesty's Stationery Office,
London, 1981.
11. Idem Chromium i n Raw and Potable Waters and
Sewage Effluents, 1980; H.M.S.O., London, 1981
12. Idem Phosphorus in Waters, Effluents and
Sewages, 1980; H.M.S.O., London, 1981.
13. Idem S i l i c o n in Waters and Effluents, 1980;
H.M.S.O., London, 1981.
14. Idem Acid-Soluble Aluminum i n Raw and Potable
Waters by Spectrophotometry H.M.S.O.
London, 1980.
15. Standard Methods For the Examination of Water
and Wastewater; American P u b l i c Health
Association: Washington, D . C . , 1985.
16. Methods for Organic Chemical Analysis of Municipal
and Industrial Wastewater; Environmental Protec-
tion Agency Publication EPA-600 14-82-057,
July 1982.
17. Statement of Work for Organics Analysis, M u l t i -
Media, Multi-Concentration, U.S. Environmental
Protection Agency, Contract Laboratory Program,
January, 1985.
18. Statement of Work for Inorganic Analysis,
Multi-Media, Multi-Concentration, U.S.
Environmental P r o t e c t i o n Agency, Contract
Laboratory Program, J u l y , 1985.
19. G l a s e r , J . A . ; Foerst, D. L . ; McKee, G. D.;
Quave, S . A . ; Budde, W.L. E n v i r o n . Sci.
Technol. 1981, 15, 1426-1435.
20. Long, G. L . ; Winefordner, J . D. A n a l . Chem.
1983, 55, 712A.
RECEIVED May 19, 1987

Chapter 5
Interlaboratory Aspects of Detection Limits

Used for Regulatory and Control
Purposes
L. B. Rogers
Department of Chemistry, University of Georgia, Athens, GA 30602
The factors that influenc

deviation for a determinatio
limit that should be used for interlaboratory
measurements are briefly reviewed. In addition,
the need for a more general approach to the estimating
and reporting of selectivity is pointed out. A
possible basis for a suitable scheme is suggested
that should be applicable to environmental and c l i n i c a l
samples that usually contain a large number of
completely unknown species, one or more of which
may interfere in the measurement of the sought-for
species.
A more n e a r l y complete p i c t u r e o f t h e problems encountered i n

i n t e r l a b o r a t o r y measurements, e s p e c i a l l y those a t t h e t r a c e l e v e l ,
may be found i n two r e c e n t publications (1,2) and i n t h e i r
references. I n c o n t r a s t , t h e p r e s e n t r e p o r t emphasizes the v a r i a b l e s
found i n i n t e r l a b o r a t o r y s t u d i e s t h a t a r e n o t a d e q u a t e l y accounted
for i n i n t r a l a b o r a t o r y s t u d i e s . I f the detection l i m i t i s c a l c u l a t e d
by t a k i n g a m u l t i p l e o f t h e r e l a t i v e s t a n d a r d d e v i a t i o n (RSD),
and i f t h e v a l u e f o r t h e p o o l o f i n t e r l a b o r a t o r y measurements i s
(as i s u s u a l l y t h e c a s e ) l a r g e r than t h a t f o r i n t r a l a b o r a t o r y
measurements, then t h e former RSD i s t h e a p p r o p r i a t e one f o r
c a l c u l a t i n g t h e d e t e c t i o n l i m i t f o r r e g u l a t o r y purposes.
This report also discusses the question of s e l e c t i v i t y ,
e s p e c i a l l y w i t h r e s p e c t t o d e t e r m i n a t i o n s o f amounts t h a t a r e near
the lower l i m i t o f d e t e c t i o n . C l a s s i c a l l y , two d i f f e r e n t , and
as n e a r l y independent procedures as p o s s i b l e , were recommended.
In t h a t c a s e , t h e minimum d e t e c t a b l e amount (MDA) would be s e t
by t h e RSD f o r t h e l e s s s e n s i t i v e p r o c e d u r e . On t h a t b a s i s , one
can conclude t h a t , f o r r e g u l a t o r y / c o n t r o l measurements i n v o l v i n g
i n t e r l a b o r a t o r y measurements o f t r a c e amounts, t h e proposed method
( d i s c u s s e d l a t e r ) f o r a r r i v i n g a t t h e A c c e p t a b l e Minimum D e t e c t a b l e
Amount (AMDA) o f t h e American N a t i o n a l Standards I n s t i t u t e (ANSI)
can be made more e x p l i c i t . Once t h a t has been done, one can proceed
0097-6156/88/0361-0094$06.00/0

5. ROGERS Interlaboratory Aspects of Detection Limits 95
to calculate the " r e g i o n s of d e t e c t a b i l i t y " and "regions of

q u a n t i t a t i o n " as recommended by a committee o f the American Chemical
S o c i e t y (_3).
However, the o r i g i n a l g o a l o f u s i n g two procedures was to
i n c r e a s e one's c o n f i d e n c e i n the r e l i a b i l i t y o f the r e s u l t through
an independent check on the s e l e c t i v i t y . In t r a c e a n a l y s e s , the
u s u a l t e s t s o f s e l e c t i v i t y , which examine the r e l a t i v e responses
from known i n t e r f e r e n c e s t h a t produce the same type o f s i g n a l ( e . g . ,
in mass s p e c t r o m e t r y , the same mass-to-charge r a t i o ) as the
s o u g h t - f o r s u b s t a n c e , appear t o be o f l i m i t e d v a l u e f o r e n v i r o n m e n t a l
and c l i n i c a l a n a l y s i s . There, one i s f a c e d w i t h the p o s s i b i l i t y
of " d i r e c t " i n t e r f e r e n c e from g e n e r a t i o n o f an i d e n t i c a l s i g n a l
from one o r more o f a v e r y l a r g e number of t r u l y unknown s p e c i e s .
T h i s s i t u a t i o n i s i n c o n t r a s t t o the more common case where the
s i g n a l o f the s o u g h t - f o r a n a l y t e i s o n l y decreased o r i n c r e a s e d
as a r e s u l t o f the presenc f th interferin species Hence
a new approach has bee
of s e l e c t i v i t y i n cases
VARIABLES IN INTERLABORATORY MEASUREMENTS

A l l o f the changes t h a t can o c c u r w i t h i n one l a b o r a t o r y can, o f
c o u r s e , o c c u r between l a b o r a t o r i e s . However, they are o f t e n much
more v i s i b l e when they o c c u r i n i n t e r l a b o r a t o r y s t u d i e s , as f o r
example, the e f f e c t s o f c o n t a m i n a t i o n from d u s t , r e a g e n t s , o r w a l l s
of c o n t a i n e r s . Two b e a u t i f u l examples from the Manhattan P r o j e c t
were r e p o r t e d t o me i n c o n v e r s a t i o n s i n 1947. In one case, C.
J . Rodden r e p o r t e d t h a t the N a t i o n a l Bureau o f Standards had g r e a t
d i f f i c u l t y i n measuring low p a r t s per m i l l i o n o f cadmium i n uranium
u n t i l they removed a l l o f the cadmium-plated ironware from t h e i r
laboratories. S i m i l a r l y , C h a r l e s Metz o f Los Alamos S c i e n t i f i c
Laboratory reported that they d i d not permit containers of
macroscopic amounts o f cadmium o r boron compounds t o be c a r r i e d
i n t o the f a c i l i t i e s where p a r t s - p e r - m i l l i o n t r a c e s o f those elements
were b e i n g a n a l y z e d i n uranium. More r e c e n t l y , Nordberg (4) r e p o r t e d
t h a t h i s attempts t o make measurements o f drugs i n body f l u i d s
l e d t o e r r a t i c , meaningless d a t a u n t i l he i s o l a t e d the l o c a t i o n
and the equipment w i t h which the c o n c e n t r a t e d doses o f drugs were
a d m i n i s t e r e d from the equipment and f a c i l i t i e s where the measurements
were made.
When a problem of c o n t a m i n a t i o n o r i n t e r f e r e n c e i s encountered
i n an i n t e r l a b o r a t o r y s t u d y , i t i s e a s i e r t o d e t e c t . Figure 1
shows the r e s u l t s o f a v e r y o l d study done by T a f t S a n i t a r y
Engineering Center, a f o r e r u n n e r o f the U. S. Environmental
P r o t e c t i o n Agency, when they sent p o r t i o n s o f a sample o f aluminum
ion t o a number o f d i f f e r e n t laboratories. Each l a b o r a t o r y
c a l i b r a t e d i t s procedure f o r aluminum i o n a g a i n s t a p r i m a r y s t a n d a r d
and d i d the a n a l y s i s i n t r i p l i c a t e . The f a c t t h a t a v a r i e t y o f
methods was used suggests t h a t the h i g h r e s u l t f o r the mean was
not due t o an i n t e r f e r e n c e i n a p a r t i c u l a r method b u t , i n s t e a d ,
due t o c o n t a m i n a t i o n o f the e n t i r e l o t o r the presence o f an
i n t e r f e r e n c e from the w a l l s of the sample c o n t a i n e r s .
Another major problem a r i s e s from l o s s e s , u s u a l l y due to
a d s o r p t i o n , but sometimes a l s o from d e c o m p o s i t i o n o r v o l a t i l i t y
of the a n a l y t e . For example, i f m i c r o p i p e t t e s are used f o r t a k i n g

2.60 TiliiII
I I I I I I I I I I I I I I I I I I I I T"| I I I I I I I I I I I 2.6^73 74 2333.10 821
ALUMINUM
AMOUNT A D 0 E 0 - I . 8 0 M 6 / L
2.40
A V E R A G E A M O U N T RECOVERED 2.21 M G / L
MEAN
2.20
2.00
O 1.80
2E
LEGEND'.
1.60
ALUM I NON
0 GRAVIMETRIC
SPECTROGRAPH
1.40 CD ERIOCHROME CYANINE R
ORTHO-PHENANTHROLINE
D NOT SPECIFIED
1.20

1.00 I I I I I II II I 1 I I I I I 1 I I 1 I I 1 II I l I I I ' I I I I i l I
R R 0
CM <M =
a : ; ; r < " x> w ro (\j o (p ro to C S I C J ( M U > I O Z * > r*. ~ ~ Z Z = Z
= CM v 5 N 1. CJ ^. - *i . ~. CM - _ ^ m c o ^ i o ^
1 0 1 0 !) K) ^ QD 0)
oo s m N N : - 1 - N r w
- - N N

LABORATORY NUMBER
F i g u r e 1. R e s u l t s o f a study by t h e T a f t S a n i t a r y E n g i n e e r i n g C e n t e r o f t h e
average v a l u e s found by d i f f e r e n t l a b o r a t o r i e s f o r samples o f t h e same unknown.
D e t a i l s are given i n text.
samples, the large surface-to-volume r a t i o may result i n r e l a t i v e l y

large losses of the sought-for species. One of the e a r l i e s t examples
of this was a report that a substantial fraction of radioactive
s i l v e r ion was lost to the walls of a micropipette even after i t
had been rinsed three times (_5). Similarly, another interlaboratory
comparison performed by Taft Sanitary Engineering Laboratory showed
that the mean found for C r ( l l l ) was substantially less than the
amount o r i g i n a l l y added to the sample (6). When one considers
that competing factors of contamination and loss are present i n
every determination, one can readily understand why r e l a t i v e standard
deviations are much larger at the trace l e v e l , especially when
carried out i n different laboratories under different environmental
conditions.
The next variable r e f l e c t s the p a r t i c i p a t i o n of more than
one person i n doing a particular determination. One of the best
examples i s found i n analyse f dioxi sample that performed
by a number of differen
amount i n a particular
was found to have a RSD of s l i g h t l y under 5%. If two people i n
that laboratory combined their e f f o r t s and did different parts
of the determination, the RSD jumped to somewhat greater than 10%.
However, when two or more people from different laboratories were
involved, the RSD value was greater than 26%. This example c l e a r l y
i l l u s t r a t e s the trend that has been confirmed i n other studies
to be mentioned l a t e r .
Another variable that leads to highly unpredictable results
i s derived from a universal sin of which we are a l l g u i l t y -
subjective use of s t a t i s t i c s , especially with respect to discarding
data. For example, i n the previous dioxin study, samples were
sent to two laboratories and the results compared. If they were
too widely discrepant, the sample was sent to a third laboratory,
and the two results that agreed the closest were reported.
Furthermore, of the results that remained, 5% were judged to be
o u t l i e r s and were discarded before calculating the 99% confidence
level of the mean. In contrast, a committee of the American Chemical
Society recommended that, i n general, o u t l i e r s should usually only
be discarded i f reasons are known why the results are bad (2)
Certainly, the dangers of discarding data are easier to see i n
other people's studies than i n one's own.
The next variable i s concerned with the experience a p a r t i c u l a r
analyst has had with a given procedure. There i s no disagreement
with the conclusion that, i n general, an experienced analyst should
have a lower RSD than one who i s inexperienced. Furthermore, one
would expect that the duration of the contact of an experienced
analyst with a p a r t i c u l a r procedure might have a noticeable influence
on the resulting RSD. However, there are two surprising results
in the data presented by Horwitz et a l . (8) which were obtained
from their examination of data from interlaboratory studies conducted
by the Association of O f f i c i a l Analytical Chemists (AOAC) (_9).
F i r s t , there was a steady decrease i n the RSD to roughly 45% of
i t s i n i t i a l value, over a period of five years! Second, at the
end of that time, although the value leveled o f f , i t was usually
between 1.5 and 3 times the value that a single analyst usually
obtained for that same procedure. It i s important to note not

98 DETECTION IN A N A L Y T I C A L C H E M I S T R Y
o n l y t h a t t h i s type o f r e s u l t was observed f o r a wide v a r i e t y o f

procedures conducted by the AOAC but a l s o t h a t i t i s c o n s i s t e n t
w i t h the t r e n d s i n the d a t a f o r d i o x i n a n a l y s e s d i s c u s s e d e a r l i e r .
The f i n a l major c o n t r i b u t i o n t o the RSD, the l a c k o f t r u l y
b l i n d q u a l i t y assurance samples, was documented i n an i n t e r l a b o r a t o r y
study b u t , undoubtedly, h o l d s as w e l l f o r a q u a l i t y assurance program
w i t h i n one laboratory. The need f o r q u a l i t y assurance samples
t h a t are " c o m p l e t e l y b l i n d " was c l e a r l y demonstrated by L i d d l e
(10) i n the f o l l o w i n g way. A group o f l a b o r a t o r i e s p a r t i c i p a t e d
i n a study i n v o l v i n g the d e t e r m i n a t i o n o f l e a d i n b l o o d serum u s i n g
unknowns d i s t r i b u t e d by L i d d l e . A l t h o u g h i t took some time f o r
the l a b o r a t o r i e s t o meet the hoped-for o v e r a l l performance level
as measured by the RSD o f the pooled r e s u l t s , when those same samples
were d i s t r i b u t e d l a t e r as complete unknowns t h a t were not suspected
of b e i n g a s s o c i a t e d w i t h the q u a l i t y assurance program, the RSD
doubled! T h i s r e s u l t does not mean t h a t the a n a l y s t s were d i s h o n e s t
Instead, i t simply r e f l e c t
care when p e r f o r m i n g a n a l y s e
e x t r a c a r e r e s u l t s i n a s i g n i f i c a n t improvement i n the r e s u l t i n g
data.
The relative contributions of the known v a r i a b l e s that
c o n t r i b u t e t o i n c r e a s e s i n the RSD f o r i n t e r l a b o r a t o r y measurements
are summarized i n Table I . The i m p o r t a n t t h i n g t o note i s t h a t ,
i n g e n e r a l , one would expect the RSD f o r q u a l i t y assurance samples
measured under t r u l y b l i n d c o n d i t i o n s t o be between 5 and 10 times
l a r g e r than those measured on known s t a n d a r d s by a s i n g l e person
in a laboratory. The r e s u l t i n g l a r g e r v a l u e f o r the RSD has an
important b e a r i n g on the l i m i t o f d e t e c t i o n f o r a procedure t h a t
i s t o be used i n i n t e r l a b o r a t o r y measurements.
Table I . V a r i a b l e s That I n c r e a s e the

Standard D e v i a t i o n
VARIABLE MULTIPLY BY
EXPERIENCE ( t i m e ) u s i n g
s t a t e d procedure (AOAC H o r w i t z ) >1
TWO OR MORE PEOPLE
Same Lab. (Seveso Study) 2
D i f f e r e n t Labs. (Seveso; AOAC) 2-4
TRULY BLIND QUALITY ASSURANCE
(Liddle-CDC) 2+
LOWER CONCS. (AMTS.) t h a t r e q u i r e
added i n d e p . c o n f i r m a t o r y measurements >1
F i n a l l y , i t i s i m p o r t a n t t o note t h a t the l i m i t o f d e t e c t i o n
w i l l be l a r g e r under some c o n d i t i o n s i n which a measurement i n v o l v e s
consumption of a sample. For example, a mass s p e c t r o m e t e r equipped
w i t h a s i n g l e d e t e c t o r i s sometimes used t o measure more than one
mass-to-charge r a t i o i n a g i v e n sample so as t o i n c r e a s e one's
c o n f i d e n c e i n the i d e n t i f i c a t i o n o f the a n a l y t e . I f the s p e c t r o m e t e r
spends e q u a l time on each o f t h r e e masses (and the s w i t c h i n g time
i s n e g l i g i b l e ) , the q u a n t i t y o f sample must be t h r e e times l a r g e r
i n o r d e r t o a t t a i n the same d e t e c t i o n l i m i t as t h a t f o r the s i m i l a r

procedure t h a t measures o n l y one mass f o r the same t o t a l time.

We s h a l l see l a t e r t h a t such a t r a d e - o f f of improved s e l e c t i v i t y
and reliability for a higher detection limit w i l l usually be
desirable.
SELECTIVITY ASPECTS
S e l e c t i v i t y i s g e n e r a l l y i n v e r s e l y r e l a t e d t o the r e l a t i v e amount
of i n t e r f e r e n c e one can expect from a p a r t i c u l a r s p e c i e s (above
a g i v e n l e v e l ) i n a t t e m p t i n g t o measure a n o t h e r s o u g h t - f o r s p e c i e s .
B e f o r e g o i n g f a r t h e r , one should r e c a l l t h a t the e x t e n t of an
i n t e r f e r e n c e i s u s u a l l y expressed i n terms of i t s c o n c e n t r a t i o n
or amount t h a t w i l l produce the same s i g n a l as the u n i t amount
of the s o u g h t - f o r species. I t i s i m p o r t a n t t o note t h a t use of
the standard a d d i t i o n or i n t e r n a l standard method t o e s t i m a t e the
amount of a s o u g h t - f o r specie provide compensatin
f o r an i n t e r f e r e n c e t h a
The e x t e n t of s i g n a l fro
taken i n t o account and s u b t r a c t e d by measuring a b l a n k t h a t c o n t a i n s
a known amount of the i n t e r f e r i n g s p e c i e s i n the presence of a l l
other s p e c i e s except the s o u g h t - f o r species. However, as will
be p o i n t e d out below, t h i s p o s s i b i l i t y i s r u l e d out i f c o m p l e t e l y
unknown substances are i n v o l v e d because one i s unable t o prepare
an a p p r o p r i a t e b l a n k .
I t i s worth n o t i n g t h a t i n c l a s s i c a l q u a n t i t a t i v e a n a l y s i s ,
which u s u a l l y d i d not i n v o l v e a n a l y s e s of t r a c e amounts, the problem
of unknown i n t e r f e r e n c e s was a t t a c k e d by u s i n g two methods t h a t
were as n e a r l y independent as p o s s i b l e . In t h a t way, the chance
t h a t an i n t e r f e r e n c e would g i v e the same response f o r each method
was m i n i m a l . In c o n t r a s t , t h e r e i s an example i n c l i n i c a l c h e m i s t r y
which c l e a r l y s u f f e r s from a l a c k of s e l e c t i v i t y i n the a c c e p t e d
method. F i g u r e 2 (11) shows t h a t the use of the determination
of sugar i n serum as a method f o r d i a g n o s i n g d i a b e t e s i s c l e a r l y
unsatisfactory. The d i s t r i b u t i o n f o r sugar c o n t e n t i n s e r a o f
d i a b e t i c s d i s t i n c t l y o v e r l a p s the d i s t r i b u t i o n f o r those who are
not. Hence, s i g n i f i c a n t f r a c t i o n s o f both f a l s e p o s i t i v e s and
f a l s e n e g a t i v e s w i l l be o b t a i n e d over a r e l a t i v e l y wide range of
sugar c o n c e n t r a t i o n . A s i m i l a r c o n c l u s i o n would have been reached
i f sugar were indeed a h i g h l y s e l e c t i v e b a s i s f o r d i a g n o s i n g d i a b e t e s
but t h e r e was a second unknown substance p r e s e n t i n some s e r a t h a t
c o n t r i b u t e d an i n t e r f e r i n g "sugar s i g n a l " . Hence, the use of two
"independent" procedures should enhance one's c o n f i d e n c e i n the
results.
The s i t u a t i o n i n v o l v i n g the completely u n i d e n t i f i e d source
of i n t e r f e r e n c e becomes i n c r e a s i n g l y i m p o r t a n t as the c o n c e n t r a t i o n
or amount of s o u g h t - f o r substance d e c r e a s e s (.2) Donaldson (12)
p o i n t e d out t h i s p r i n c i p l e , which can be i l l u s t r a t e d u s i n g 99.999%
water t h a t c o n t a i n s 10 p a r t s per m i l l i o n o f t o t a l i m p u r i t y . If
we make the s i m p l i f y i n g assumption t h a t a l l i m p u r i t i e s are p r e s e n t
at the same l e v e l , we c a l c u l a t e t h a t 10 i m p u r i t i e s can be p r e s e n t
at the 1 ppm l e v e l , 10^ a t the 1 p a r t per b i l l i o n l e v e l or 10^
at 1 p a r t per t r i l l i o n . The l a s t f i g u r e c o r r e s p o n d s q u i t e c l o s e l y
t o the e s t i m a t e d t o t a l of known c h e m i c a l s p e c i e s . Hence, when
one i s faced w i t h the t r a c e a n a l y s i s of a complete unknown, such

s
50 1
1 r i ' I
40
30, DISTRIBUTION OF T H O S E
WITHOUT 0 I A 8 E T E S
o 20
OC
Ui
CD
2 10
z 3
- DISTRIBUTION OF T H O S E
WITH D I A B E T E S 2
>

40 80 120 160 200 240 280 320 360 >
BLOOD SUGAR - mtj/ lOOcc
5
Figure 2. D i s t r i b u t i o n of blood sugar i n d i a b e t i c s and nondiabetics.

as the water i n a drainage ditch from an i n d u s t r i a l plant, the

approach to s e l e c t i v i t y of estimating the extent of interference
based on a few known species i s t o t a l l y inadequate.
One can get a feeling for the r e a l i t y of such interferences
from unusual examples that came to l i g h t that relate to my service
with a team of consultants to advise the Dow Chemical Company
concerning the adequacy of their procedures for analyzing dioxins.
The overall procedure used by Dow involved multiple high-resolution
steps (13); an i n i t i a l l i q u i d - l i q u i d extraction followed successively
by l i q u i d chromatography, gas chromatography, and, f i n a l l y , a few
selected ions i n mass spectrometry. One of their chemists remarked
that they found, quite by accident, when using highly p u r i f i e d
nitrogen to evaporate a solvent from a sample p r i o r to mass
spectrometric measurement of the dioxin, high e r r a t i c values were
obtained unless they f i r s t passed the nitrogen through a column
packed with s i l i c a g e l That seemingly unnecessary step removed
the unidentified interference
attention only recently
fish that had been extracted, and "cleaned up" by liquid
chromatography before being analyzed by mass spectrometry using
three different mass-to-charge r a t i o s , unexpectedly high results
were obtained. By recording the entire spectrum over a wide range
of masses, the selected masses were found to be s i t t i n g on top
of a very broad high background which the chemist speculated might
have been a large glyceride, unanticipated and unidentified, that
broke up into a broad continuum of different masses i n the range
of interest. (Telephone conversation with Lewis Shadoff i n July
1986)
A third example from the laboratory of V e i l l o n has recently
been described, together with i t s h i s t o r i c a l background (14). In
b r i e f the graphite furnace measurements of chromium using atomic
absorption spectrometry, were confounded with an unidentified species
in urine samples that was contributing a background signal which
was inadequately corrected for when using a deuterium lamp source.
V e i l l o n showed a linear plot r e l a t i n g the magnitude of the background
signal to the apparent chromium concentration i n the urine. He
estimated that erroneous values had been reported i n the l i t e r a t u r e
for more than 10 years!
E a r l i e r , mention was made of the c l a s s i c a l way of insuring
higher s e l e c t i v i t y by doing two unrelated determinations i n p a r a l l e l .
The Dow determinations of dioxins cited above are examples i n which
selectivity was increased by performing multiple, relatively
high-resolution steps i n one quantitative procedure, but they s t i l l
encountered unexpected interferences. There are also other examples
of procedures consisting of multiple high-resolution steps such
as those in mass spectrometry-mass spectrometry, mass
spectrometry-infrared, and other "hyphenated" techniques which
also increase one's confidence i n the r e s u l t s . However, i t i s
clear that we need a way to estimate the s e l e c t i v i t y of such
techniques when a large number of species, most of them unknown,
might be present. This aspect w i l l be addressed more f u l l y l a t e r .

REGULATORY/CONTROL LIMITS FOR TRACE AMOUNTS
The term " r e g u l a t o r y l i m i t " u s u a l l y i m p l i e s t h a t a governmental

body has s e t t h e l i m i t whereas t h e term " c o n t r o l l i m i t " i s u s u a l l y
associated with process specifications f o r private or
non-governmental purposes. I t i s the w r i t e r ' s impression t h a t ,
i n a r r i v i n g a t such l i m i t s , t h e u s u a l p r a c t i c e i s t o : ( a ) use t h e
most s e n s i t i v e p r o c e d u r e , ( b ) t e s t t h e procedure w i t h known o r
l i k e l y i n t e r f e r e n c e s , ( c ) use d a t a from i n t r a l a b o r a t o r y s t u d i e s
t o e s t i m a t e t h e RSD f o r use i n c a l c u l a t i n g t h e d e t e c t i o n limit
f o r the p r o c e d u r e , (d) use a q u a l i t y a s s u r a n c e procedure t h a t employs
samples o f unknown c o n c e n t r a t i o n s w h i c h , however, a r e known t o
the a n a l y s t t o be s t a n d a r d s , and (e) a p p l y g u i d e l i n e s o f the American
Chemical S o c i e t y i n s e l e c t i n g the " r e g i o n s o f d e t e c t i o n " and " r e g i o n s
of q u a n t i t a t i o n " (_3) .
In c o n t r a s t , a g e n e r a l p r o t o c o l recommended b t h AOAC that
has long been used by t h
from t h e above approach
s e l e c t e d f o r use may o r may n o t be the most s e n s i t i v e one a v a i l a b l e .
Second, t h e RSD o b t a i n e d from t h e pooled r e s u l t s has been r e c o g n i z e d
as u s u a l l y b e i n g a p p r o x i m a t e l y double t h a t o f t h e v a l u e obtained
in a single laboratory. The f a c t t h a t a l o n g time p e r i o d may be
r e q u i r e d t o reach t h a t minimum m u l t i p l e o f an i n t r a l a b o r a t o r y v a l u e
has a l s o been r e c o g n i z e d .
The American N a t i o n a l Standards I n s t i t u t e (ANSI), a f t e r much
d e l i b e r a t i o n , has added a n o t h e r c o n c e p t , t h a t o f an A c c e p t a b l e
Minimum D e t e c t i o n Amount (AMDA) f o r use i n i n t e r l a b o r a t o r y s t u d i e s .
A l t h o u g h i t has been d i s c u s s e d i n more d e t a i l by ( B r o d s k y , A. ,
N u c l e a r R e g u l a t o r y Commission, Washington, D.C., u n p u b l i s h e d d a t a . ) ,
the AMDA, i n b r i e f , i s t h e MDA t h a t i s o b t a i n a b l e i n p r a c t i c e by
competent a n a l y t i c a l s c i e n t i s t s u s i n g the best state-of-the-art
at e c o n o m i c a l l y a f f o r d a b l e c o s t s t o t h e customer. Where p o s s i b l e ,
the AMDA has been e s t a b l i s h e d a t l e a s t s e v e r a l times h i g h e r than
the MDA, i f t h e h i g h e r v a l u e i s adequate f o r p r o t e c t i o n purposes.
In essence, t h e AMDA p e r m i t s the f l e x i b i l i t y for different
l a b o r a t o r i e s t o use cheaper and more r a p i d procedures when they
are a v a i l a b l e and meet t h e l i m i t . In the w r i t e r ' s opinion, t h i s
concept has been i m p l i c i t , i f n o t e x p l i c i t , i n t h e s e l e c t i o n o f
procedures by the AOAC and i n EPA r e g u l a t i o n s .
The AMDA concept o f ANSI i s a t t r a c t i v e , and t h e w r i t e r would
l i k e t o expand upon i t . F i r s t , i f t h e RSD and t h e MDA have been
based on i n t r a l a b o r a t o r y s t u d i e s , a procedure f o r e s t i m a t i n g t h e
i n t e r l a b o r a t o r y RSD and t h e AMDA c o u l d be done by a p p l y i n g an
appropriate s e r i e s o f approximate c o r r e c t i o n s shown i n T a b l e 1.
Such c o r r e c t i o n s c o u l d be a p p l i e d t o e i t h e r t h e o r i g i n a l procedure
f o r which an AMDA was t o be s t a t e d o r t o the a l t e r n a t e l e s s s e n s i t i v e
procedure which was t o be used f o r t h e AMDA. Of c o u r s e , i f an
AOAC-type i n t e r l a b o r a t o r y study were conducted t o determine a pooled
RSD, o n l y the c o r r e c t i o n s f o r v a r i a b l e s n o t i n c l u d e d i n the AOAC-type
study would be used.
To a p p l y t h e AMDA concept t o c r i t i c a l r e g u l a t o r y d e c i s i o n s ,
i t would be easy t o adopt t h e c l a s s i c a l concept o f r e q u i r i n g a n a l y s e s
by two as n e a r l y independent methods as p o s s i b l e . T h i s would be
an i m p o r t a n t s t a r t i n t h e d i r e c t i o n o f a d d r e s s i n g t h e q u e s t i o n s
of p o s s i b l e c o n t r i b u t i o n s t o t h e measured s i g n a l s by unsuspected

and unrecognized interferences. However, since the second procedure

w i l l r a r e l y be as sensitive as the f i r s t , adoption of the second
procedure w i l l provide a basis for deciding more e x p l i c i t l y the
minimum multiple that should be applied by ANSI i n c a l c u l a t i n g
the detection l i m i t on going from an MDA to the AMDA (or to a second,
confirmatory AMDA procedure).
The use of such a double-check w i l l also contribute to the
r e l i a b i l i t y through having another test of the s e l e c t i v i t y . Remember
that the usual way of discussing s e l e c t i v i t y and interferences
i s to work with species that are known or suspected to be present.
However desirable that exercise may be, i t i s , unfortunately,
u n r e a l i s t i c i n dealing with most environmental samples and, possibly,
many c l i n i c a l samples. If one r e c a l l s the large number of possible
interferences that might be present when making an analysis at
the low part per b i l l i o n l e v e l , a new approach to thinking about
possible interferences and estimating s e l e c t i v i t y of a procedure
appears to be needed.
hopeless, undefined proble
of Kaiser (15).
Let us assume that one can do a computerized search of physical
properties of a l l known organic and inorganic substances so that
one can, for example, assemble l i s t s of their molecular weights,
b o i l i n g points, mass-to-charge r a t i o s of positive (or negative)
molecular ions, major bands in the infrared, and so forth. Then,
i t should be possible to plot a d i s t r i b u t i o n of the frequencies
with which d i f f e r e n t values are found along the axis of each
property. Taking the b o i l i n g point at atmospheric pressure i n
degrees Kelvin as an example of one property, i t would then be
possible to estimate f i r s t the number of compounds that had b o i l i n g
points between 300K and 350K (or any other desired range) and,
second, what f r a c t i o n of the t o t a l possible number of species to
which that corresponded. Hence, i f a d i s t i l l a t i o n were performed
over that range, one would be making an estimate of the number
of possible compounds that might be present. If a second property
were employed i n the o v e r a l l analysis, such as a positive molecular
ion spectrum in mass spectrometry, a similar estimate could be
made of that d i s t r i b u t i o n . (For example, Harvan et a l . (16) reported
that there were 700 elemental compositions within 0.2 dalton of
the mass for tetrachlorodioxin that s a t i s f i e d the c r i t e r i a for
valence, contained four chlorine atoms and, i n addition, only C,H,N,0
and S atoms.) It should then be possible to calculate for the
n_ properties involved in a given o v e r a l l procedure a "volume" i n
n-dimensional space which would provide an estimate of the number
of possible interferences present. That number, when viewed i n
terms of the t o t a l number of compounds considered could provide
a basis for estimating the s e l e c t i v i t y of the o v e r a l l procedure.
To c i t e a s p e c i f i c example, a computerized search of the current
Heilbron compilation, published by Chapman and H a l l , was made using
the Lockheed Dialog system. That f i l e currently includes 150,000
organic compounds. F i r s t , a boiling-point range between 200 and
220C was examined, and a f t e r less than 5 seconds of search time,
1321 compounds were reported as having been found. Second, molecular
weight was searched between 300 and 305. Again, a f t e r less than
5 seconds of search, 1492 compounds were found. F i n a l l y , a search

for the compounds found i n both l i s t s took less than 2 seconds

to come up with a t o t a l of 6. Hence, when suitable properties
are available and the software permits, appropriate searches can
be made very quickly.
It i s important to note that i m p l i c i t assumptions are made
in doing such a calculation. The f i r s t i s that the compounds i n
the c o l l e c t i o n are representative of the entire body of organic
compounds. If so. one can make the following extrapolation. If
there are 1.5 x 10 compounds estimated to e x i s t , one can extrapolate
11 7
to 60 possible " h i t s ; i f 1.5 x 10 compounds are assumed, 600
" h i t s " are estimated. If further assumptions are made that (a)
the i n t e r v a l i s small r e l a t i v e to the width of the o v e r a l l
d i s t r i b u t i o n and (b) the d i s t r i b u t i o n of b o i l i n g points i s uniform
across the 20 i n t e r v a l , then, i f one can j u s t i f i a b l y narrow the
range of b o i l i n g points to 10C instead of 20, the number of
7
estimated " h i t s " could b reduced t 300 t f th 1.5 10
compounds. Clearly, assumption
t o t a l number of compound
property parameter w i l l have i t s own c h a r a c t e r i s i t i c a b i l i t y to
impart s e l e c t i v i t y , especially when combined with one or more other
parameters, the use of additional parameters should produce
significant reductions in the estimated number of "hits".
Unfortunately, the number of compounds that can be searched for
specific infrared bands, mass-to-charge ratios of ions, gas
chromatographic retention indices, and l i q u i d chromatographic indices
i s much smaller than 150,000 species. However, one can expect
the data bases for a l l types of compounds - organic, inorganic,
and metal lorganic - to grow quickly during the next few years.
Although one can foresee disagreements about the t o t a l to be used
for the number of possible species, this should not be a major
deterrent to the use of this approach for estimating r e l a t i v e
s e l e c t i v i t i e s of different properties and the gains to be made
from an improved precision of measurements. (For example, many
oligomeric series, e. g., those of polystyrene terminated with
butyl, as well as the corresponding series terminated with other
a l k y l groups, w i l l quickly add many species to the estimated t o t a l .
Improvements i n precision and accuracy with which properties can
be measured w i l l usually conteract that increase by reducing the
percentage of " h i t s " . However, i n special cases, such as i n high
resolution mass spectrometry, of large organic compounds, where
isotopic species can e a s i l y double or treble the number of major
molecular ions that should be detectable within 5 millimass units,
greater resolution does not necessarily insure a better qualitative
or quantitative r e s u l t . (17A)) In any case, the approximate approach
to the estimation of o v e r a l l s e l e c t i v i t y of the type outlined here
should be useful to the analyst for comparing two o v e r a l l procedures.
In a l i g h t e r vein, i t would be of interest for a chemist to
find out, for example, the mean value for the b o i l i n g point of
a l l chemicals (ruling out those that decompose as well as a few
extremes l i k e l i q u i d helium and magnesium oxide). Similarly, i t
would be interesting to speculate about the mean value i n the
d i s t r i b u t i o n of v i b r a t i o n a l frequencies or of negative molecular
ions.
Before concluding this discussion of s e l e c t i v i t y , one must
r e a l i z e that there are situations i n which the analyst might l i k e

to have two methods that exhibited as nearly the same s e l e c t i v i t y

as possible for a wide variety of compounds. That s e l e c t i v i t y
goal, which represents the opposite extreme to the e a r l i e r one,
has been represented schematically (19) i n Figure 3 which shows
the scopes of d i f f e r e n t procedures for determining the organic
carbon content of water. Instead, to a greater or lesser extent
that depends on the selections made, the reported carbon content
i s a function of the types of compounds present and the procedure
employed for the determination.
CONCLUSIONS
F i r s t , i t i s clear that any stated minimum detection l i m i t for

a procedure used for interlaboratory measurements should incorporate
the RSD that includes a l l known sources of interlaboratory
uncertainty. Second, because
the sought-for substanc
for minimizing the e f f e c t get larger
i t would be highly desirable that, i n c r i t i c a l regulatory decisions,
a minimum of two as nearly independent procedures as possible be
used. In that case, the less sensitive procedure would be the
one that would have to be used to establish the AMDA. At that
point, one would then apply the recommended c r i t e r i a of the American
Chemical Society Committee to calculate values for the "region
of detection" and "region of quantitation" (3).
Third, a basis has been presented for a general approach for
estimating and reporting s e l e c t i v i t y for a sample i n which a large
number of unsuspected and unknown interferences may occur. The
approach depends upon n-dimensional screens of the properties of
the sample i t s e l f (e.g., s o l u b i l i t y ) and of those involved i n the
i s o l a t i o n and measurement procedures.
F i n a l l y , i t i s important to note that this discussion has
focussed almost e n t i r e l y on uncertainties a r i s i n g largely from
chemical sources. These always assume p r i o r c a l i b r a t i o n against
a primary chemical standard and, frequently, additional internal
references, except in c l i n i c a l chemistry where the rule has not
yet been universally adopted (20). However, f a i l u r e s to control
the laboratory environment (21) and to calibrate instruments properly
(22 and correspondence from G. N. Bowers, J r . on May 15, 1984.)
(e.g., wavelength, detector response, instrument and room
temperatures, flow rate, resistance between electrodes) and computer
algorithms (23-27) (e.g., peak deconvolution, curve-fitting,
multivariate analysis) are also major sources of uncertainty and
error. It seems clear that one must not assume that the use of
costly complex instrumentation, including those incorporating on-line
computers for instrument control, data a c q u i s i t i o n and data analysis,
eliminates the need for careful environmental controls, frequent
c a l i b r a t i o n of the c r i t i c a l components of the instrumentation against
primary standards, and c a l i b r a t i o n (validation) of the algorithms
for analyses of data.

KEY A r e a h-Sulfite, nitrite

A r e a i - N o n b i o d g r a d a b l organics
F i g u r e 3. Schematic diagram o f m a t e r i a l s measured by c h e m i c a l

oxygen demand (COD) as r e l a t e d t o t o t a l o r g a n i c m a t e r i a l s .
(Reproduced w i t h p e r m i s s i o n from Ref. 19. C o p y r i g h t 1973
American Water Works A s s o c i a t i o n . )

ACKNOWLEDGMENT
The w r i t e r wishes t o thank the U. S. Department o f Energy, Department

of B a s i c E n e r g i e s , C o n t r a c t No. DE-AS09-76ER00854, f o r l o n g - t e r m
support o f h i s r e s e a r c h which has l e d t o awareness o f t h e problems
d i s c u s s e d i n t h i s paper.
LITERATURE CITED
1. Rogers, L. B. J. Chem. Ed. 1986, 63, 3.
2. Ad Hoc Subcommittee Dealing with the Scientific Aspects of
Regulatory Measurements, ACS Joint Board/Council Committee
on Science, Improving the R e l i a b i l i t y and Acceptability of
Analytical Chemical Data used for Public Purposes May 10,
1982; see also Chem. Eng. News 1982, 60(23), 44.
3. ACS Committee on Environmental Improvement Principles of
Environmental Analysi
MacDougall, D.; Crummett
4. Nordberg, R. The Sampling Situation and the Analytical Result
Symposium on Trace Analysis of Drugs and Related Compounds
in Complex Mixtures, Stockholm, Sweden, November 18, 1981;
Abstracted in Acta Pharm. Suecica 1982, 19(1), 51; see also:
Borg, O., Idem. 52.
5. Hershenson, H. M.; and Rogers, L. B. Anal. Chem. 1952, 24,
219; see also, Rogers, L. B. J. Chem. Ed. 1952, 29, 612.
6. Data shown in Figure 4 of reference 1.
7. di Domencio, A . ; Merli, F.; Boniforti, L.; Camoni, I . ; Di
Muccio, A . ; Taggi, F . ; Vergori, L.; Colli, G.; Elli, G.; Gorni,
A.; Grassi, P.; Invernizzi, G.; Jemma, A . ; Luciani, L . ;
Cattabeni, F . ; De Angelis, L . ; G a l l i , G.; Chiabrando, C.;
Faneli, R. Anal. Chem. 1979, 51, 735..
8. Horwitz, W.; Kamps, L. R.; Boyer, K. W. J. Assoc. Off. Anal.
Chem. 1980, 63, 1344.
9. Garfield, F. M. Quality Assurance Principles for Analytical
Laboratories, Association of Official Analytical Chemists,
Arlington, VA, 1984.
10. Liddle, J . A . ; through Maugh, T. H . , II Science, 1982, 215,
490.
11. Weinstein, M. C. Med. Decis. Making 1981, 1, 309 through Lusted,
L. B. Recent Advances in Analytical Methodology in the Life
Sciences, L. A. Beaver, Ed.; U. S. Food and Drug Administration,
Office of Science Coordination, Washington, D.C., 1982; p.
13.
12. Donaldson, W. T. Environ. Sci. Technol. 1977, 11, 348-351.
13. Nestrick, T. J.; Lamparski, L. L . ; Stehl, R. H. Anal. Chem.
1979, 51, 1453, 2273.; see also Hummell, R. A . ; Shadoff, L.
A. Anal. Chem. 1980, 52, 191.
14. Veillon, C. Anal. Chem. 1986, 58, 851A; see also Guthrie,
B. E . ; Wolf, W. R.; Veillon, C. Anal. Chem. 1978, 50, 1900.
15. Kaiser, H. Spectrochim. Acta 1978, Part B, 33B, 551; see also
Trehy, M. L.; Yost, R. A . ; Dorsey, J . G. Anal. Chem. 1986,
58, 14.
16. Harvan, D. J.; Hass, J . R.; Schroeder, J . L.; Corbett, B.
J. Anal. Chem., 1981, 53, 1755..

17. Kateman, K; Pijpers, F. W. In Chemical Analysis; Elving, P.J.;

Winefordner, J . D.: Wiley, New York, 1981, Vol 60, p.155.
17A. Mahle, N. H.; Shadoff, L. H. Biomed. Mass Spectrom. 1982,
9, 45; see Table 9 and the related discussion.
18. Analytical Reference Service Training Program, Report of Water
Metals, No. 2, U. S. Dept. of Health, Education and Welfare,
Robert A. Taft Sanitary Engineering Center, Cincinnati, OH,
1962.
19. Stevens, A. A; Symons, J . M. Proc. Amer. Water Works Assoc.
Water Quality Technology Conference, 1973, pp. xxiii-1 to
xxiii-26.
20. Bowers, G. N., Jr.; McComb, R. B. Clin. Chem, 1984, 30, 1128.
21. de Haseth, J . A. Appl. Spectrosc. 1982, 36, 544.
22. Horwitz, W. J . Assoc. Off. Anal. Chem. 1984, 67, 1053.
23. Meglen, R. R. Practical Consideration for Acceptance of Pattern
Recognition and Othe Chemometri Method 100th AOAC Annual
Meeting, Scottsdale
24. Sprouse, J . F. Spectroscopy, , 1(6),
25. Gritton, V. Calibration Curves for Flame AAS - A Study No.
123, Div. Anal. Chem., Natl. Mtg. Amer. Chem. Soc., Anaheim,
CA, Sept. 11, 1986.
26. Marshall, A. G.; Chen, L . ; Cottrell, C. E. Limits to Precision
in Measurement of Peak Height, Position, and Width in Fourier
Transform Spectrometry No. 67, Div. Anal. Chem., Natl. Mtg.
Am. Chem. Soc., Anaheim, CA, Sept. 10, 1986.
27. Rothman, L. D. Chromatogr. Forum 1986, 1(2), 13.
RECEIVED November 13, 1986

Chapter 6
Comparison of Detection Limits

in Atomic Spectroscopic Methods
of Analysis
Michael S. Epstein
Inorganic Analytical Research Division, National Bureau of Standards,

The compariso
fundamental part of many decision-making processes for
the analytical chemist. Despite numerous efforts to
standardize methodology for the calculation and
reporting of detection limits, there is still a wide
divergence in the way they appear i n the literature.
This paper discusses valid and invalid methods to
calculate, report, and compare detection limits using
atomic spectroscopic techniques. Noises which limit
detection are discussed for analytical methods such as
plasma emission spectroscopy, atomic absorption
spectroscopy and laser excited atomic fluorescence
spectroscopy.
The c o m p a r i s o n o f d e t e c t i o n l i m i t s i s a fundamental p a r t o f most

decision-making processes f o r the a n a l y t i c a l chemist. whether t h e
d e c i s i o n i n v o l v e s t h e purchase o f a new i n s t r u m e n t o r t h e d e s i g n o f
a t r a c e a n a l y s i s p r o t o c o l , t h e f i g u r e - o f - m e r i t [1] w h i c h i n f l u e n c e s
the c h o i c e w i l l most l i k e l y be t h e d e t e c t i o n l i m i t . S i n c e one o r
more o f t h e t e c h n i q u e s b e i n g compared i s o f t e n u n f a m i l i a r , t h e
d e c i s i o n w i l l be b a s e d on i n f o r m a t i o n t h a t can be r e t r i e v e d from t h e
literature, both from m a n u f a c t u r e r a d v e r t i s i n g and t h e open
scientific literature. Unfortunately, despite the e f f o r t s of
o r g a n i z a t i o n s such as t h e I n t e r n a t i o n a l U n i o n o f Pure and A p p l i e d
C h e m i s t r y (IUPAC) t o s t a n d a r d i z e methodology t o c a l c u l a t e and r e p o r t
d e t e c t i o n l i m i t s [ 2 ] , t h e r e i s s t i l l a wide d i v e r g e n c e i n t h e way
t h a t d e t e c t i o n l i m i t s appear i n p r i n t . W h i l e t h e r e i s h o p e f u l l y no
d e l i b e r a t e attempt on t h e p a r t o f a u t h o r s and m a n u f a c t u r e r s t o b i a s
d e t e c t i o n l i m i t s towards a p a r t i c u l a r technique, t h e manner o f
calculating and r e p o r t i n g c a n l e a d t o a m i s i n t e r p r e t a t i o n o f
d e t e c t i o n l i m i t s by the c a r e l e s s o r u n f a m i l i a r reader. I f the
d e t e c t i o n l i m i t methodology i s n o t w e l l documented, a c o m p a r i s o n c a n
be b i a s e d by s e v e r a l o r d e r s o f magnitude.
I t i s impossible t o completely eliminate bias i n detection
limit comparisons, particularly when comparing detection
capabilities i n r e a l sample m a t r i c e s . However, i f t h e b a s i c
This chapter not subject to U.S. copyright

Published 1988 American Chemical Society

p r i n c i p l e s b e h i n d t h e t e c h n i q u e s t o be compared a r e u n d e r s t o o d and
we a r e aware o f t h e common ways i n w h i c h d e t e c t i o n l i m i t comparisons
can be m i s i n t e r p r e t e d , r e a s o n a b l y v a l i d c o n c l u s i o n s c a n be drawn.
Thus, t h i s d i s c u s s i o n w i l l c o n c e n t r a t e i n g e n e r a l on v a l i d and
i n v a l i d ways t o compare d e t e c t i o n l i m i t s and i n s p e c i f i c d e t a i l
about l i m i t i n g n o i s e s w h i c h determine those d e t e c t i o n l i m i t s u s i n g
several o f t h e most common atomic spectroscopic techniques,
i n c l u d i n g flame and plasma e m i s s i o n s p e c t r o s c o p y , atomic a b s o r p t i o n
s p e c t r o s c o p y , and l a s e r - e x i t e d atomic f l u o r e s c e n c e s p e c t r o s c o p y .
B e f o r e d e t e c t i o n l i m i t s a r e d i s c u s s e d i n any d e t a i l , i t i s
n e c e s s a r y t o d e f i n e t h e scope o f t h e p r o c e s s t o w h i c h t h e d e t e c t i o n
l i m i t a p p l i e s . F o r example, t h e d e t e c t i o n l i m i t d e t e r m i n e d f o r an
element i n t h e absence o f c o n c o m i t a n t s (i.e., i n pure water
s o l u t i o n ) i s l i k e l y t o be s i g n i f i c a n t l y l e s s t h a n t h e d e t e c t i o n
l i m i t d e t e r m i n e d f o r a complete a n a l y t i c a l p r o t o c o l w h i c h i n c l u d e s
s a m p l i n g , sample p r e p a r a t i o n and a n a l y s i s The former which i s
the type o f d e t e c t i o n l i m i
may be r e f e r r e d t o a
i n s t r u m e n t a l n o i s e sources w h i c h a r e i n h e r e n t i n t h e a n a l y t i c a l
i n s t r u m e n t used. Fundamental d e t e c t i o n l i m i t s a r e o f t e n o f l i m i t e d
v a l u e t o t h e p r a c t i c i n g a n a l y t i c a l chemist who must determine t h a t
element i n r e a l and o f t e n v e r y complex m a t r i c e s . The l a t t e r type o f
d e t e c t i o n l i m i t , r e f l e c t i n g t h e e n t i r e a n a l y t i c a l p r o t o c o l , may be
r e f e r r e d t o as m e t h o d o l o g i c a l . Methodological d e t e c t i o n l i m i t s are
a l s o o f l i m i t e d v a l u e s i n c e they i n c l u d e many v a r i a b l e s w h i c h cannot
be e a s i l y reproduced. The d e t e c t i o n l i m i t s t o be d i s c u s s e d here
w i l l be c a l l e d i n s t r u m e n t a l and w i l l be d e f i n e d as f a l l i n g between
fundamental and m e t h o d o l o g i c a l i n t h a t they w i l l c o n s i d e r v a r i a t i o n s
i n d u c e d by t h e i n s t r u m e n t a l o n e and by t h e i n t e r a c t i o n o f t h e sample
w i t h the instrument, but w i l l not consider the e n t i r e a n a l y t i c a l
scheme w h i c h i n c l u d e s b l u n d e r s and c o n t a m i n a t i o n i n t h e s a m p l i n g and
sample p r e p a r a t i o n p r o c e s s . I t i s noteworthy t h a t i n s t r u m e n t a l
d e t e c t i o n l i m i t s w i l l approach fundamental d e t e c t i o n l i m i t s when t h e
sample m a t r i x i s s i m p l e o r when n o i s e r e d u c t i o n methods s p e c i f i c t o
sample-matrix-induced noises are a p p l i e d .
While the d i s c u s s i o n w i l l d e a l w i t h atomic rather than
m o l e c u l a r s p e c t r o s c o p i c methods, many o f t h e p o i n t s t o be made w i l l
a p p l y t o b o t h atomic and m o l e c u l a r methods. The major d i f f e r e n c e
between t h e n o i s e c h a r a c t e r i s t i c s o f t h e two methods i s u s u a l l y t h e
dynamic o r f l o w i n g s t a t e o f an atomic system, such as a h i g h
temperature flame o r plasma, compared t o t h e s t a t i c s t a t e o f a
m o l e c u l a r system i n w h i c h t h e sample u s u a l l y i s p l a c e d i n a s m a l l
transparent cuvette. The dynamic s t a t e o f t h e atomic system
generates an a n a l y t e s i g n a l - c a r r i e d n o i s e w h i c h i s p r o p o r t i o n a l t o
a n a l y t e s i g n a l magnitude and thus becomes l i m i t i n g a t h i g h a n a l y t e
concentrations. (A s i g n a l - c a r r i e d n o i s e i s one whose magnitude i s a
c o n s t a n t p e r c e n t a g e o f t h e a m p l i t u d e o f a s i g n a l , w h i c h may be due
to background o r t o t h e a n a l y t e . Thus, an a n a l y t e s i g n a l - c a r r i e d
n o i s e i s a f l u c t u a t i o n i n t h e phenomenon caused by t h e a n a l y t e ,
where t h e phenomenon i s used as a measure o f t h e a n a l y t e
c o n c e n t r a t i o n , such as a b s o r p t i o n o r e m i s s i o n o f e l e c t r o m a g n e t i c
radiation). The s t a t i c s t a t e o f t h e m o l e c u l a r system l i m i t s t h e
magnitude o f a n a l y t e s i g n a l - c a r r i e d n o i s e s , e x c e p t where t h e s t a t i c
s t a t e i s d i s t u r b e d ( i . e . , v i b r a t i o n , c e l l p o s i t i o n changes, e t c . ) o r
where r a d i a t i o n source f l u c t u a t i o n s a r e s i g n i f i c a n t a t h i g h a n a l y t e
concentrations ( i . e . , molecular fluorescence spectrophotometry).

6. EPSTEIN Comparing Detection Limits 111
B i a s In D e t e c t i o n L i m i t Comparisons
There a r e s e v e r a l ways t h a t d e t e c t i o n l i m i t i n f o r m a t i o n can be

presented i n o r d e r t o b i a s the o b s e r v e r . Again, i t must be
emphasized t h a t i n most cases s u f f i c i e n t i n f o r m a t i o n w i l l be
p r e s e n t e d i n a f i g u r e or i n the accompanying t e x t t o a l l o w the
knowledgeable reader to properly interpret detection limit
comparisons.
R e a l or A r t i f i c i a l D e t e c t i o n L i m i t s . C e r t a i n l y , one o f the most

common ways t o r e p o r t d e t e c t i o n l i m i t s i s i n "pure aqueous
s o l u t i o n . " whether the a n a l y t i c a l c o n d i t i o n s or the i n s t r u m e n t a t i o n
used i s c a p a b l e o f those d e t e c t i o n l i m i t s when r e a l samples are
analyzed i s another question. T h i s source o f b i a s i s most o f t e n
e n c o u n t e r e d when a new a n a l y t i c a l technique i s developed. An
example i s the e a r l development f flam atomi fluorescenc
spectroscopy (FAFS), wher
consumption b u r n e r i
C e r t a i n l y no one would attempt r e a l sample a n a l y s i s i n such a flame
because o f i t s t u r b u l e n t f l o w and poor d i s s o c i a t i o n c h a r a c t e r i s t i c s .
More r e a l i s t i c d e t e c t i o n l i m i t s are on the o r d e r o f 200 pg/mL i n an
a i r - a c e t y l e n e flame [4] . I n flame atomic a b s o r p t i o n s p e c t r o s c o p y
(FAAS), t i n d e t e c t i o n l i m i t s are s i g n i f i c a n t l y b e t t e r (4x) in a
cool air-hydrogen flame t h a n i n h o t t e r flames as a r e s u l t o f
i n c r e a s e d s e n s i t i v i t y and lower flame background e m i s s i o n [ 5 ] . The
use o f the s a m p l i n g b o a t [6] i n FAAS a l s o improves d e t e c t i o n l i m i t s
f o r many elements by an o r d e r o f magnitude because o f i n c r e a s e d
sample t r a n s p o r t e f f i c i e n c y . However, n e i t h e r o f these t e c h n i q u e s
i s w i d e l y used i n FAAS, s i n c e b o t h e x h i b i t s i g n i f i c a n t chemical
i n t e r f e r e n c e s w i t h r e a l samples. Recently developed techniques,
such as i n d u c t i v e l y - c o u p l e d plasma mass s p e c t r o m e t r y (ICP-MS) [7]
and l a s e r enhanced i o n i z a t i o n s p e c t r o s c o p y (LEIS) [8] e x h i b i t
s i m i l a r sample-related degradation of d e t e c t i o n l i m i t s . C e r t a i n l y ,
most t e c h n i q u e d e t e c t i o n l i m i t s s u f f e r somewhat when r e a l samples
are a n a l y z e d and n o i s e s induced by the sample m a t r i x become
limiting. The e x t e n t o f t h i s e f f e c t w i l l v a r y , however, from
t e c h n i q u e t o t e c h n i q u e , and w i l l u s u a l l y d i m i n i s h as the method
reaches m a t u r i t y .
Detection Limit Criterion. The c r i t e r i o n used t o d e f i n e the

d e t e c t i o n l i m i t , or perhaps as i m p o r t a n t , the p r o t o c o l used t o
measure i t , can be c r i t i c a l i n e s t a b l i s h i n g a v a l i d d e t e c t i o n l i m i t .
C u r r i e [9] has d e s c r i b e d the wide v a r i a t i o n i n d e t e c t i o n l i m i t
definitions for radiochemical measurements reported in the
literature. IUPAC [2] recommends the d e t e c t i o n l i m i t , c , L be
defined as the concentration of an analyte equal to a
background-corrected s i g n a l , x-^ - xg, t h r e e times the estimated
s t a n d a r d d e v i a t i o n o f a s i n g l e d e t e r m i n a t i o n u s i n g 20 measurements
o f the b l a n k .
X = X k s
L B + B W
c L = ks /m B (2)
where

x L uncorrected signal
xg blank measure
sg - estimated standard deviation of the blank measure
c L * detection l i m i t , which i s the concentration
derived from the smallest measure ( x ) that can L
be detected with reasonable confidence.

k - numerical factor chosen i n accordance with the
confidence l e v e l desired.
m - analytical sensitivity
As pointed out by Long and Winefordner [10], the use of k-3 allows a
confidence l e v e l of 99.86% for a normal d i s t r i b u t i o n of xg, or an
89% confidence l e v e l for a non-normal d i s t r i b u t i o n . While xg w i l l
often be normally d i s t r i b u t e d when instrumental noise limits
detection, the presence of analyte contamination i n the blank,
either i n the sample preparation process or as a series of discrete
events ( i . e . , Na or Fe
measurement process, w i l
a d i s t r i b u t i o n may be bimodal or skewed depending on the source and
c h a r a c t e r i s t i c s of the contaminant. Long and Winefordner [10] have
also presented several examples of the influence of measurement
protocol on c-^. The use of values of k < 3 or the use of the
standard deviation of the mean or pooled standard deviation rather
than the standard deviation of a single measurement, can lead to C L
values which deviate by an order of magnitude from the IUPAC model.
Measurement protocols which include the error i n the a n a l y t i c a l
s e n s i t i v i t y as well as the error i n the blank can also cause C L to
deviate s i g n i f i c a n t l y from the IUPAC model, which assumes a
well-defined s e n s i t i v i t y . F i n a l l y , the presence of very low
frequency noise or d r i f t may not be incorporated into the IUPAC
d e f i n i t i o n of detection l i m i t [11]. The c a l i b r a t i o n scheme used for
r e a l samples may be spread out over a longer time period than was
used for the determination of the detection l i m i t and thus noises
which were i n s i g n i f i c a n t during the detection l i m i t measurement may
be encountered. Ideally, a technique detection l i m i t should be
determined using the measurement protocol employed for r e a l sample
analysis.
A n a l y t i c a l Blank. I f the detection l i m i t i s not measured from the

true a n a l y t i c a l blank, a c r i t i c a l part of the detection l i m i t
determination has been ignored. Since the emphasis i n this
discussion i s on "instrumental" rather than "methodological"
detection l i m i t s , only blanks r e s u l t i n g from the instrumentation
w i l l be considered. Although method blanks can c e r t a i n l y be
l i m i t i n g , p a r t i c u l a r l y for elements such as Fe, Na, and Ca which are
common i n the laboratory environment, they are not as predictable as
instrumental contamination blanks, since variations i n laboratory
procedures and design w i l l be much greater than variations i n
instrument design. For example, i n FAFS one can s i g n i f i c a n t l y
improve a detection l i m i t i n a s i t u a t i o n l i m i t e d by flame scatter of
source r a d i a t i o n by making measurements with no water being
introduced into the flame. By eliminating the scattering species,
unvaporized water droplets, the detection l i m i t i s improved.
S i m i l a r l y , one can measure a graphite furnace atomic absorption

s p e c t r o s c o p y (GFAAS) d e t e c t i o n l i m i t w i t h o u t a c t u a l l y a t o m i z i n g a
b l a n k sample, assuming the n o i s e t o be independent o f the a t o m i z e r .
T h i s i s c e r t a i n l y an i n v a l i d assumption when d e t e r m i n i n g an element
whose most s e n s i t i v e a b s o r p t i o n l i n e l i e s i n the v i s i b l e r e g i o n o f
the spectrum, such as barium, where t h e r m a l e m i s s i o n from the
g r a p h i t e tube i s s i g n i f i c a n t , or where c o n t a m i n a t i o n i n the tube i s
l i m i t i n g , such as when z i n c i s determined. For t h e s e elements,
p u b l i s h e d d e t e c t i o n l i m i t s may be i n v a l i d , u n l e s s t h e y were measured
under a c t u a l a n a l y s i s c o n d i t i o n s . The m o r a l i s thus t o measure the
b l a n k under c o n d i t i o n s as s i m i l a r as p o s s i b l e t o the a n a l y s i s
c o n d i t i o n s used.
Instrument N o i s e C h a r a c t e r i s t i c s . Depending on the f r e q u e n c y domain

spectrum o f the s i g n a l from the a n a l y t i c a l i n s t r u m e n t , t h a t i s , i f
the n o i s e i s w h i t e ( s h o t ) o r 1/f ( f l i c k e r ) i n n a t u r e [ 1 ] , the
i n t e g r a t i o n time o r tim constant d fo th detectio limit
d e t e r m i n a t i o n may have a
I n cases where s h o t n o i s
a l l wavelengths o r FAAS above 230 nm, o r ICP e m i s s i o n s p e c t r o s c o p y
(ICP-ES) below 250 nm, the d e t e c t i o n l i m i t can be improved as the
s q u a r e - r o o t o f the i n t e g r a t i o n time. In f l i c k e r noise l i m i t e d
c a s e s , t h e r e may be l i t t l e or no improvement i n the d e t e c t i o n l i m i t
w i t h an i n c r e a s e i n the i n t e g r a t i o n time. [12,13] The improvement
i n d e t e c t i o n l i m i t f o r an i n c r e a s e o f i n t e g r a t i o n time w i l l be
u l t i m a t e l y l i m i t e d by the s i g n i f i c a n c e o f v e r y low f r e q u e n c y d r i f t
and the a v a i l a b i l i t y o f l a r g e volumes o f sample s o l u t i o n .
Measurement U n i t s . Perhaps the most o b v i o u s y e t c o n f u s i n g a s p e c t o f

many d e t e c t i o n l i m i t comparisons i s the use o f " r e l a t i v e " v e r s u s
"absolute" u n i t s . R e l a t i v e u n i t s r e f l e c t a mass p e r u n i t volume,
such as micrograms per m i l l i l i t e r , w h i l e a b s o l u t e u n i t s r e f l e c t a
mass o n l y , such as micrograms. O b v i o u s l y , " r e l a t i v e " and " a b s o l u t e "
u n i t s s h o u l d n o t be d i r e c t l y compared. However, a b s o l u t e u n i t s can
be c o n v e r t e d i n t o r e l a t i v e u n i t s and v i c e v e r s a , employing the
volume o f s o l u t i o n u t i l i z e d by a p a r t i c u l a r t e c h n i q u e . Nonetheless,
how that conversion i s done or how i t i s documented can
s i g n i f i c a n t l y b i a s the o b s e r v e r . Table I i l l u s t r a t e s s e v e r a l
examples, t a k e n from the s c i e n t i f i c l i t e r a t u r e , o f the use o f
d e t e c t i o n l i m i t v a l u e s i n a t a b l e f o r comparison purposes. I n each
case the a u t h o r p r o v i d e s adequate i n f o r m a t i o n f o r the i n f o r m e d
reader t o make an accurate comparison. Nevertheless, the
c o n c l u s i o n s drawn by the c a r e l e s s o r u n i n f o r m e d r e a d e r who does not
r e a d o r u n d e r s t a n d the f o o t n o t e s or the t e x t w h i c h d e s c r i b e s the
t a b l e , can be b i a s e d by s e v e r a l o r d e r s o f magnitude.
T a b l e l a , p r e s e n t s a comparison o f FAAS and GFAAS d e t e c t i o n
l i m i t s [14]. Without r e a d i n g the t e x t which r e f e r s t o the t a b l e ,
one i s impressed by the s i g n i f i c a n t , 3 t o 4 o r d e r s o f magnitude,
improvement u s i n g the g r a p h i t e f u r n a c e . However, the t e x t c l e a r l y
i n d i c a t e s t h a t b o t h flame and g r a p h i t e f u r n a c e d e t e c t i o n l i m i t s
assume a 1 mL sample volume which, w h i l e c e r t a i n l y v a l i d i n a flame,
i s n o t v a l i d f o r a g r a p h i t e f u r n a c e s i n c e the maximum sample volume
i s u s u a l l y about 50 t o 100 /xL. Some systems, l i k e the c a r b o n r o d
a t o m i z e r [ 1 5 ] , can o n l y accommodate 1 t o 2 /iL o f s o l u t i o n . Thus,
f o r a v a l i d comparison, the g r a p h i t e f u r n a c e d e t e c t i o n l i m i t s must
be degraded by one t o two o r d e r s o f magnitude.

Table l a . D e t e c t i o n L i m i t s R e p o r t e d f o r Atomic A b s o r p t i o n
D e t e c t i o n L i m i t (zxg/mL)
Element Flame Nonflame
Ba 0.02 6 x 10" 6
Ca 0.002 4 x 10" 7
Fe 0.004 1 x 10" 5
Mn 0.0008 2 x 10" 7
"...the l i m i t s are based on a s i g n a l - t o - n o i s e r a t i o = 2 c r i t e r i o n and

the assumption t h a t a volume o f 1 mL i s the minimum r e q u i r e d f o r a
d e t e r m i n a t i o n . For example, i f an a b s o l u t e d e t e c t i o n l i m i t i s g i v e n
( e . g . , nonflame a t o m i z e r ) as 10"^ g, t h i s i s e x p r e s s e d as a
c o n c e n t r a t i o n a l d e t e c t i o n l i m i t o f 0.001 /xg/mL. One must b e a r i n
mind t h a t most c u r r e n t nonflame a t o m i z e r s cannot h a n d l e samples
l a r g e r t h a n , say, 0.1 mL
cannot h a n d l e samples ( f o
mL. The 1 mL c r i t e r i o n . .. i s thus more f o r the purpose o f d i r e c t
comparison than f o r the v e r y l o w e s t p o s s i b l e d e t e c t i o n l i m i t s . . . "
R e p r i n t e d from [14] by p e r m i s s i o n o f J o h n W i l e y and Sons, c o p y r i g h t
1976.
Source: Reproduced w i t h p e r m i s s i o n from Ref. 14. Copyright 1976
Wiley.
I n T a b l e l b [16] a b s o l u t e r a t h e r t h a n r e l a t i v e d e t e c t i o n l i m i t s
are compared f o r s e v e r a l t e c h n i q u e s . U n l e s s one l o o k s a t the
f o o t n o t e s however, i t i s not obvious t h a t the d e t e c t i o n l i m i t f o r
one method i s based on a 1 /xL sample s i z e , a n o t h e r on a 5 /iL sample
s i z e and another on a 1 mL sample s i z e , thus b i a s i n g the c a r e l e s s
o b s e r v e r o f t h i s t a b l e by 2 t o 3 o r d e r s o f magnitude.
Table l b . Absolute D e t e c t i o n L i m i t s
U s i n g Atomic F l u o r e s c e n c e S p e c t r o m e t r y and S e v e r a l Other Methods
D e t e c t i o n L i m i t s (pg)
Element AFS AAS AEICP
Ag 0.4 0.2 200
Cd 0.0015 0.1 70
Mg 1 0.06 3
Ni 5 10 200
AFS = Atomic f l u o r e s c e n c e s p e c t r o m e t r y - 1 /xL sample s i z e

AAS = Atomic a b s o r p t i o n s p e c t r o m e t r y - 5 /xL sample s i z e
AEICP = Plasma e m i s s i o n u s i n g the ICP - 1 mL sample s i z e [16]
R e p r i n t e d from [16] by p e r m i s s i o n o f Pergamon J o u r n a l s L t d .
Source: Reproduced w i t h p e r m i s s i o n from Ref. 16. C o p y r i g h t 1979
Pergamon P r e s s .
F i n a l l y , i n Table l c [17] a comparison i s made of d e t e c t i o n
l i m i t s f o r carbon f u r n a c e atomic e m i s s i o n s p e c t r o s c o p y (CFAES),
flame e m i s s i o n s p e c t r o s c o p y ( F E S ) , and CFAAS. Note t h a t
s e n s i t i v i t i e s are used as p s e u d o - d e t e c t i o n l i m i t s f o r CFAAS. These
are not r e a l l y s e n s i t i v i t i e s as d e f i n e d by IUPAC [ 1 8 ] , but are
c h a r a c t e r i s t i c c o n c e n t r a t i o n s , s i n c e they r e p r e s e n t a c o n c e n t r a t i o n
e q u i v a l e n t to an absorbance of 0.0044. Furthermore, n o i s e

measurements are n o t made i n the c a l c u l a t i o n o f t h i s parameter, so

the t r u e d e t e c t i o n l i m i t w i l l l i k e l y be much s m a l l e r , p a r t i c u l a r l y
in the case o f CFAAS, where t r a n s m i s s i o n flicker noise is
negligible.
T a b l e l c . D e t e c t i o n L i m i t s U s i n g Carbon Furnace
Atomic E m i s s i o n S p e c t r o m e t r y and Other Techniques
D e t e c t i o n L i m i t s (ttg/mL)
Element CFAES Flame e m i s s i o n CFAAS
Mo 0.016 0.03 0.005

Si 0.088 10 0.01
Be 0.46 10 0.0002
CFAES = Carbon f u r n a c e atomi
of s o l u t i o n
CFAAS = Carbon f u r n a c e atomi a b s o r p t i o spectrometry y
i n /xg/mL/0.0044 A - based on a 20 fih a l i q u o t o f s o l u t i o n
R e p r i n t e d from [17] by p e r m i s s i o n o f E l s e v i e r S c i e n c e P u b l i s h e r s
Source: Reproduced w i t h p e r m i s s i o n from Ref. 17. Copyright 1978
Elsevier Scientific.
Now, l e t us summarize the q u e s t i o n s t h a t s h o u l d be c o n s i d e r e d
when comparing d e t e c t i o n l i m i t s .
F i r s t , what i s the n o i s e bandwidth ( d e f i n e d by the i n t e g r a t i o n
time o r time c o n s t a n t f o r each measurement) o f each instrument?
Were the measurements made under s i m i l a r c o n d i t i o n s and, when u s i n g
a method such as ICP-ES [12,13], does i t make any d i f f e r e n c e ?
Second,' are we d e a l i n g w i t h a b s o l u t e o r r e l a t i v e u n i t s and have
the u n i t s been c o r r e c t l y c o n v e r t e d t o a l l o w a v a l i d comparison?
T h i r d , does sample-induced n o i s e , t h a t i s n o i s e r e s u l t i n g from
components i n the sample m a t r i x , s i g n i f i c a n t l y degrade d e t e c t i o n
limits? T h i s may be more s i g n i f i c a n t f o r some t e c h n i q u e s than
others. For example, s c a t t e r or m o l e c u l a r a b s o r p t i o n i n FAAS, when
compensated f o r by a background c o r r e c t i o n method such as Zeeman
s p l i t t i n g o r a continuum s o u r c e , w i l l u s u a l l y r e s u l t i n o n l y a s m a l l
i n c r e a s e i n shot n o i s e due to a t t e n u a t i o n of primary source
i n t e n s i t y and no s i g n i f i c a n t change i n d e t e c t i o n l i m i t s w i l l o c c u r .
The same m a t r i x components i n an ICP-ES system, w h i c h i s f l i c k e r
n o i s e l i m i t e d , may show a f a r more s i g n i f i c a n t d e g r a d a t i o n of
d e t e c t i o n l i m i t when f l i c k e r i n the sample m a t r i x e m i s s i o n becomes
the l i m i t i n g n o i s e .
F o u r t h , does the s e n s i t i v i t y of the t e c h n i q u e d e c r e a s e i n the
p r e s e n c e of the sample m a t r i x ? O f t e n c o n d i t i o n s w h i c h f a v o r the
b e s t d e t e c t i o n l i m i t s , such as low b a c k g r o u n d e m i s s i o n or h i g h
sample i n t r o d u c t i o n r a t e s a l s o r e s u l t i n r e d u c e d sample d i s s o c i a t i o n
and t h u s d e c r e a s e d a n a l y t e s e n s i t i v i t y when a complex sample m a t r i x
i s present. Are d e t e c t i o n l i m i t s d e t e r m i n e d under u n r e a l i s t i c
c o n d i t i o n s or w i t h a p p a r a t u s u n s u i t a b l e f o r r e a l sample a n a l y s i s ?
F i f t h , a r e we d e a l i n g w i t h c o n d i t i o n s o p t i m i z e d f o r a s i n g l e
element or m u l t i e l e m e n t a n a l y s i s ? Compromise c o n d i t i o n s degrade
d e t e c t i o n l i m i t s b u t improve the i n f o r m i n g power of the method
( i . e . , the t o t a l amount of i n f o r m a t i o n about a sample t h a t can be
o b t a i n e d from an a n a l y t i c a l method).

Sixth and f i n a l l y , what c r i t e r i a were used to define the

detection l i m i t and how was i t calculated?
Noises Which Limit Detection
Let us now look b r i e f l y into the three major classes of a n a l y t i c a l

spectrometric methods: emission, absorption, and fluorescence.
Noises w i l l be defined, and examples of how and when they l i m i t
detection w i l l be given. Table II l i s t s the major noises which
l i m i t detection for the three atomic spectroscopic techniques to be
discussed. Detailed d e f i n i t i o n s of these noises may be found i n the
paper by Epstein and Winefordner [1].
TABLE I I . Noises Which Limit Detection

i n Atomic Spectroscopic Methods
PMT shot noise induced by dark current, atomizer background

emission, or sample matrix emission.
Electronics noise (including RF)
Atomizer background intensity fluctuations induced by atomizer
gases, sample matrix components, or contamination.
ABSORPTION
PMT shot noise induced by the radiation source, atomizer background

emission, or sample matrix emission.
Electronics noise
Radiation source intensity fluctuations
Atomizer transmission fluctuations induced by flame or furnace
gases, sample matrix components, or contamination.
FLUORESCENCE
PMT shot noise induced by dark current, atomizer background

emission, sample matrix emission, or scattered r a d i a t i o n source
intensity.
Electronics noise (including RF)
Radiation source intensity fluctuations carried by scatter,
contamination fluorescence, or broadband fluorescence from flame and
furnace gases or from sample matrix components.
Atomizer background intensity fluctuations induced by flame or
furnace gases, sample matrix components, or contamination.
Every spectrometric system consists of four of the components

shown i n Figure 1: (a) a source of atoms; (b) a spectrometer to
isolate the radiation whose intensity and frequency contains
information about the analyte; (c) a photodetector to convert
photons to e l e c t r i c current; and (d) a signal processing scheme to

R ADI A T I O N
SOURCE
(Fluorescence)
RADIATION ATOMIC OPTICAL

S IGNAL
SOURCE SPECTROMETER
VAPOR CELL DETECTOR PROCESSING
(Absorption)
NON-OPTICAL
DETECTION

Figure 1. Components of an a n a l y t i c a l spectrometric system.

decode the i n f o r m a t i o n s t o r e d i n the r a d i a t i o n . In emission

methods, the h i g h temperature o f the atom s o u r c e p r o v i d e s e x c i t a t i o n
energy t o promote e l e c t r o n t r a n s i t i o n s , w h i l e i n a b s o r p t i o n and
f l u o r e s c e n c e methods, e x t e r n a l r a d i a t i o n s o u r c e s a r e u s e d t o induce
the e l e c t r o n t r a n s i t i o n s o f the a n a l y t e atoms. A l l o f these
i n s t r u m e n t components can r e s u l t i n n o i s e w h i c h l i m i t s d e t e c t i o n .
S p e c t r o s c o p i c t e c h n i q u e s which use n o n - o p t i c a l d e t e c t i o n , such as
LEIS o r p h o t o a c o u s t i c s p e c t r o s c o p y , a r e c h a r a c t e r i z e d by n o i s e
s o u r c e s s i m i l a r t o f l u o r e s c e n c e , s i n c e the i n f o r m a t i o n - c a r r y i n g
phenomenon i s energy r e l e a s e f o l l o w i n g a b s o r p t i o n . R a t h e r than
r a d i a t i o n a l d e a c t i v a t i o n o f the e x c i t e d s t a t e , i n LEIS the energy
release mechanism i s flame ion current g e n e r a t i o n , and in
photoacoustic spectroscopy, it is thermal or collisional
deactivation.
E m i s s i o n N o i s e Sources N o i s e s i n the e m i s s i o n t e c h n i q u e a r e the

s i m p l e s t to understand
d i r e c t c u r r e n t plasma (DCP
become p o p u l a r i n r e c e n t y e a r s . The n o i s e s w h i c h l i m i t d e t e c t i o n
u s i n g t h e s e e m i s s i o n s o u r c e s are e a s i l y c h a r a c t e r i z e d . W i t h v e r y
low o p t i c a l throughput, such as when narrow s l i t w i d t h s a r e used i n
the f a r UV, p h o t o m u l t i p l i e r dark c u r r e n t n o i s e may be s i g n i f i c a n t .
However, i n most c a s e s , s h o t n o i s e i n d u c e d by the s o u r c e background
r a d i a t i o n , o r f l i c k e r n o i s e c a r r i e d by the s o u r c e background are
l i m i t i n g . The background i n t e n s i t y may r e s u l t from argon e m i s s i o n i n
the s o u r c e o r may be i n d u c e d by i n t e r a c t i o n o f the s o u r c e w i t h the
sample m a t r i x . I n the case o f f l i c k e r n o i s e , t h a t i s , the
f l u c t u a t i o n i n the background i n t e n s i t y , the n o i s e u s u a l l y r e s u l t s
from t e m p o r a l v a r i a t i o n s i n the sample t r a n s p o r t system o r the
e x t e r n a l gas f l o w s .
The major q u e s t i o n when comparing d e t e c t i o n l i m i t s u s i n g
e m i s s i o n t e c h n i q u e s i s whether the s i g n a l - t o - b a c k g r o u n d r a t i o (SBR)
o r the s i g n a l - t o - n o i s e r a t i o (SNR) was used as the measure o f
d e t e c t i o n l i m i t . The SBR r e q u i r e s a measure o f the c o n c e n t r a t i o n
c o r r e s p o n d i n g t o a m u l t i p l e o f the background i n t e n s i t y , r a t h e r than
the n o i s e , and thus r e q u i r e s o n l y one measurement o f background.
The measurement o f background i s u s u a l l y made i n the presence o f the
a n a l y t e - c o n t a i n i n g sample by measuring a t a w a v e l e n g t h slightly
o f f s e t from the wavelength o f the a n a l y t e i n t e n s i t y maximum. I n a
m u l t i e l e m e n t system, i t i s thus much s i m p l e r t o m o n i t o r i n s t r u m e n t
performance by measuring the SBR f o r each c h a n n e l , r a t h e r t h a n the
SNR, w h i c h would r e q u i r e m u l t i p l e measurements. The d e t e c t i o n l i m i t
i s t h e n c a l c u l a t e d by assuming t h a t the b a c k g r o u n d - c a r r i e d f l i c k e r
n o i s e i s l i m i t i n g and t h a t t h e r e i s a c o n s t a n t r e l a t i v e s t a n d a r d
d e v i a t i o n o f the background e m i s s i o n , u s u a l l y about one p e r c e n t .
The l i m i t a t i o n t o t h i s procedure has been c l e a r l y p o i n t e d out by
Boumans [ 1 9 ] , who d e s c r i b e s the r e l a t i o n s h i p o f the relative
s t a n d a r d d e v i a t i o n o f the background t o the f l i c k e r n o i s e and shot
n o i s e components by the f o l l o w i n g e q u a t i o n :
2 2
(RSD) B = (a B + g/Vx )^
B O)
where
(RSD) B = observed r e l a t i v e s t a n d a r d d e v i a t i o n o f the
background e m i s s i o n .

ag = f l i c k e r f a c t o r i n d u c e d by v a r i a t i o n s i n
i n s t r u m e n t a l components such as n e b u l i z e r o r gas
flow controls.
xg measured background s i g n a l i n u n i t s o f anode
current.
g - photomultiplier gain
f$ c o n s t a n t c o e f f i c i e n t w h i c h i n c l u d e s components due
to the e f f e c t i v e system n o i s e b a n d w i d t h , the
e l e c t r o n i c charge, and g a i n f l u c t u a t i o n s due t o
secondary e l e c t r o n e m i s s i o n .
The v a l i d i t y o f assuming a c o n s t a n t s t a n d a r d d e v i a t i o n o f the

background e m i s s i o n depends on the dominance o f background f l i c k e r
noise. W i t h t h a t n o i s e , w h i c h Boumans p o i n t s out i s l i m i t i n g a t
w a v e l e n g t h s g r e a t e r t h a n 300 nm, the SNR and thus the d e t e c t i o n
l i m i t can be c h a r a c t e r i z e d b th SBR d i v i d e d b flicke factor
ag the f i r s t term i
f u n c t i o n of a p a r t i c u l a
depends on the s t a b i l i t y o f v a r i o u s i n s t r u m e n t components. Thus, as
l o n g as f l i c k e r n o i s e i s l i m i t i n g and the f l i c k e r f a c t o r does not
change, the a p p r o x i m a t i o n i s v a l i d . D e v i a t i o n s from the a s s u m p t i o n
occur a t s h o r t e r w a v e l e n g t h s , where the spectrometer optical
t h r o u g h p u t and plasma background i n t e n s i t y d e c r e a s e . The background
s h o t n o i s e i n t e n s i t y , r e p r e s e n t e d by the second term i n Boumans'
e q u a t i o n , ( g ^ / x g ) ^ , w i l l make a s i g n i f i c a n t c o n t r i b u t i o n t o the
v a r i a t i o n o f the background i n t e n s i t y , and the s i m p l e r e l a t i o n s h i p
o f f l i c k e r f a c t o r t o SNR mentioned p r e v i o u s l y b r e a k s down. Thus,
d e t e c t i o n l i m i t s c a l c u l a t e d from SBR's w i t h o u t c o n s i d e r a t i o n o f shot
n o i s e may be i n e r r o r .
A b s o r p t i o n N o i s e Sources. N o i s e s i n atomic a b s o r p t i o n s p e c t r o s c o p y
a r e more complex than i n e m i s s i o n . When a source o f r a d i a t i o n i s
i n t r o d u c e d , whose a t t e n u a t i o n c a r r i e s the a n a l y t e i n f o r m a t i o n ,
s e v e r a l new l i m i t i n g n o i s e s o u r c e s are i n t r o d u c e d . F l i c k e r noise
due t o e m i s s i o n from the h i g h temperature atomic v a p o r c e l l i s not
as s i g n i f i c a n t as i t i s i n e m i s s i o n t e c h n i q u e s , because atomic
a b s o r p t i o n uses source m o d u l a t i o n t o d i s c r i m i n a t e a g a i n s t such n o i s e
by encoding the a n a l y t e i n f o r m a t i o n s i g n a l a t a h i g h frequency.
Shot n o i s e i s s t i l l observed as a r e s u l t o f background e m i s s i o n from
the flame o r from sample m a t r i x components, b u t no s i g n i f i c a n t
f l i c k e r n o i s e i s measured. However, new n o i s e s a r e s h o t and f l i c k e r
from the r a d i a t i o n s o u r c e , flame t r a n s m i s s i o n f l i c k e r n o i s e w h i c h
becomes l i m i t i n g a t wavelengths l e s s t h a n 230 nm, and m o l e c u l a r
a b s o r p t i o n o r s c a t t e r n o i s e from sample m a t r i x components.
A l l o f the f l i c k e r n o i s e s can be e f f e c t i v e l y e l i m i n a t e d by the
use of double-beam o p t i c s i n c o n j u n c t i o n w i t h a background
c o r r e c t i o n system such as Zeeman s p l i t t i n g or a w e l l - a l i g n e d ( o r
wavelength-modulated) continuum s o u r c e . Thus the u l t i m a t e l i m i t i n g
n o i s e i n atomic a b s o r p t i o n i s source shot n o i s e , w h i c h can be
r e d u c e d ( r e l a t i v e t o t o t a l source i n t e n s i t y o r I ) by i n c r e a s i n g the
Q
source i n t e n s i t y , up t o the p o i n t o f o p t i c a l s a t u r a t i o n .
Table I I I presents some examples o f l i m i t i n g n o i s e s in
d i f f e r e n t atomic a b s o r p t i o n d e t e r m i n a t i o n s . These measurements are
a c o m p i l a t i o n o f i n f o r m a t i o n from s e v e r a l s o u r c e s , b u t p r i m a r i l y
from the work o f I n g l e [20,21] u s i n g a v e r y s i m p l e , single-beam

atomic absorption spectrometer. Aluminum in a nitrous

oxide-acetylene flame i s l i m i t e d by flame transmission f l i c k e r and
source f l i c k e r . The flame transmission f l i c k e r r e s u l t s from
absorption by molecular species (e.g., OH). Barium i n a similar
flame, but with a double-beam instrument, i s l i m i t e d only by source
induced shot noise. The double-beam system reduces the source
f l i c k e r noise component. Calcium i n an air-acetylene flame i s
l i m i t e d by both source f l i c k e r and shot noise, while i n the hotter
nitrous oxide-acetylene flame, flame emission shot noise becomes
greater than the source shot noise. The flame emission shot noise
results from the intense molecular emissions of CN and CH i n the
higher temperature flame. Copper i s l i m i t e d by both source f l i c k e r
and shot noise at a one second integration time, but l i m i t e d by only
f l i c k e r noise at a ten second integration time, a r e s u l t of the
reduction of the shot noise component. An increase i n integration
time w i l l improve the detection l i m i t i n a shot noise l i m i t e d case
Table I I I . Dominant
Element Wavelength (nm) Flame type Limiting noise

a b
(Absorbance<0.2)
Al 309.3 N2O/C2H2 Flame transmission

flicker
Radiation source f l i c k e r
c
Ba 553.6 N2O/C2H2 Radiation source-induced
shot noise
Ca 422.6 Air/C2H2 Radiation source f l i c k e r

Radiation source- induced
shot noise
Ca 422.6 N2O/C2H2 Radiation source f l i c k e r

Flame background emission
noise
d
Cu ( l s ) 324.7 Air/C2H2 Radiation source f l i c k e r
Radiation source-induced
shot noise
d
Cu ( 1 0 s ) 324.7 Air/C2H2 Radiation source f l i c k e r
Zn 213.8 Air/C2H2 Flame transmission

flicker
a
Determined with a single-beam AAS instrument [20,21], except as
noted.
^Noises with a variance of at least 33% of the most s i g n i f i c a n t
noise.
c
Determined with a double-beam AAS instrument.
Întegration time

F i n a l l y , zinc i s an example of an element whose absorption

wavelength i s lower than 230 nm, the range where flame transmission
noise dominates. Let us look a b i t more c l o s e l y at the case of
zinc. The instrument performance can be characterized using a
p r e c i s i o n p l o t , shown i n Figure 2, where the r e l a t i v e standard
deviation of concentration i s p l o t t e d on the v e r t i c a l axis and
concentration on the h o r i z o n t a l axis. The detection l i m i t i s
defined by the i n t e r s e c t i o n of the p r e c i s i o n curve with the RSD
which represents the c r i t e r i o n used to define detection, about 30%
RSD for k - 3. Note that flame transmission f l i c k e r noise l i m i t s
detection.
I f the burner head i s rotated to reduce s e n s i t i v i t y , we f i n d
that the l i m i t i n g noise i s no longer flame transmission f l i c k e r , but
source shot noise, since the absorption path has been reduced by a
factor of 20. Although the s e n s i t i v i t y i s decreased by a factor of
20, the detection l i m i t i decreased b onl facto f 10 sinc
the flame transmission nois
back to a statement mad
c h a r a c t e r i s t i c concentration [18] cannot be used as an accurate
measure of detection l i m i t i n AAS. Unlike the case of SBR i n
emission, because of the complexity of noises i n atomic absorption,
a general and simple relationship cannot be derived to relate
c h a r a c t e r i s t i c concentration and detection l i m i t .
Laser Fluorescence Noise Sources. F i n a l l y , l e t us examine a

technique with very complex noise c h a r a c t e r i s t i c s , laser excited
flame atomic fluorescence spectrometry (LEAFS). In t h i s technique,
not only are we dealing with a radiation source as well as an atomic
vapor c e l l , as i n atomic absorption, but the source i s pulsed with
pulse widths of nanoseconds to microseconds, so that we must deal
with very large incident source photon fluxes which may r e s u l t i n
o p t i c a l saturation, and very small average signals from the atomic
vapor c e l l at the detection l i m i t [22]. Detection schemes involve
gated amplifiers, which are synchronized to the laser pulse incident
on the flame and which average the analyte fluorescence pulses [23].
The l i m i t i n g noises can vary s i g n i f i c a n t l y depending on the
configuration of the o p t i c a l system, the type of flame, and the
type, temporal pulse width, and i n t e n s i t y of the laser used.
Electronic noise, dark current noise, and flame background shot and
f l i c k e r noise are temporally continuous noises that tend to l i m i t
laser-based spectroscopic systems whose detection system gatewidths
are i n the microsecond range and wider. At smaller detector
gatewidths, the temporally continuous noises w i l l be exceeded by
pulse-type noises which occur only during the laser pulse and thus
during the open gate of the detection system. These are
laser-induced noises such as laser scatter f l i c k e r and shot noise,
laser induced nonanalyte fluorescence f l i c k e r and shot noise, and
radiofrequency (RF) noise.
If we look at a few cases i n the l i t e r a t u r e , we see a wide
v a r i a t i o n i n the reported l i m i t i n g noises [24-27]. Table IV l i s t s
noises and detection l i m i t s for several laser systems. A
flashlamp-pumped system, because of the r e l a t i v e l y low i n t e n s i t y and
wide temporal pulse width, about 1 /zs, i s l i m i t e d by flame emission
shot and f l i c k e r , which are temporally continuous noises. The

ioo r flame transmission

flicker noise limited
_ radiation source shot noise
\
\ limited
10
limited
Q
CO |
analyte absorption flicker noise limited
0.1 _L_ _1_ J_ l

0.01 0.1 I 10 100 1000 10000
Zinc (jjg/mL)
F i g u r e 2. P r e c i s i o n p l o t f o r t h e d e t e r m i n a t i o n o f z i n c by FAAS
u s i n g a measurement c e l l (burner head) whose l o n g a x i s (10 cm) i s
p a r a l l e l (--) o r p e r p e n d i c u l a r (-X-) t o t h e o p t i c a l p a t h o f t h e
spectrometer.

h i g h e r i n t e n s i t y l a s e r s , w i t h v e r y narrow t e m p o r a l w i d t h s , on the
o r d e r o f 5-to-20 n s , a r e l i m i t e d by p u l s e type n o i s e s such as RF o r
s c a t t e r shot and f l i c k e r . D e t e c t i o n l i m i t s f o r these systems c a n be
c o r r e l a t e d q u i t e w e l l w i t h t h e i n t e n s i t y , t e m p o r a l and s p e c t r a l
p u l s e w i d t h , r e p e t i t i o n r a t e , and e x c i t a t i o n a r e a o f t h e l a s e r [ 2 7 ] .
The l a s e r most l i k e l y t o p r o v i d e improved d e t e c t i o n l i m i t s i s t h e
copper vapor l a s e r , s i n c e i t i s n o t l i m i t e d by a fundamental n o i s e ,
such as shot n o i s e , b u t r a t h e r by RF, w h i c h c a n be r e d u c e d by p r o p e r
shielding.
TABLE IV. L i m i t i n g N o i s e s and D e t e c t i o n L i m i t s

i n L a s e r - E x c i t e d Atomic F l u o r e s c e n c e S p e c t r o m e t r y
u s i n g the 296.7/373.5 nm T r a n s i t i o n o f I r o n
Detection Limit
Dye L a s e r Pump Majo
Flashlamp Flame e m i s s i o n s h o t / f l i c k e r 0.6 [25]
N 2 laser RF, b o x c a r 30 [24]
Excimer l a s e r Contamination, boxcar, RF, 0.2 [26]
Nd:YAG l a s e r
Focused Scatter shot/flicker 7 [27]
Beam expanded Flame e m i s s i o n s h o t 36 [27]
Cu v a p o r l a s e r RF 0.8 [27]
d e t e c t i o n l i m i t s assume a 1 s time c o n s t a n t and SNR - 3.
Conclusions
F i n a l l y , a few s u g g e s t i o n s s h o u l d be made f o r d e t e r m i n i n g
detection limits.
Whatever c r i t e r i o n and p r o t o c o l a r e used s h o u l d be r e p o r t e d i n
detail. When s u f f i c i e n t i n f o r m a t i o n i s p r o v i d e d , t h e r e a d e r can
normalize the reported values to other detection limit
methodologies, p a r t i c u l a r l y i f he i s f a m i l i a r w i t h t h e n o i s e
c h a r a c t e r i s t i c s o f the methods compared.
Determine the analyte signal or response (i.e., the
s e n s i t i v i t y ) w i t h an a n a l y t e c o n c e n t r a t i o n c l o s e enough t o t h e b l a n k
n o i s e l e v e l t o a s s u r e t h a t l i n e a r i t y o f response e x i s t s down t o t h e
noise l e v e l . Do n o t assume t h a t v a r i a t i o n o f a n a l y t e response
equates w i t h t h a t o f the b l a n k response.
Measure the v a r i a t i o n o f the blank under instrumental
c o n d i t i o n s and w i t h the measurement p r o t o c o l t y p i c a l l y used f o r
a n a l y s i s o f r e a l samples. E s t a b l i s h t h e e f f e c t o f r e a l sample
m a t r i c e s on t h e s e n s i t i v i t y and n o i s e l e v e l o f t h e i n s t r u m e n t when
o p e r a t e d under these c o n d i t i o n s .
I t i s c e r t a i n that despite a l l precautions, the v a l i d i t y of

detection limit comparisons w i l l still remain i n question.

Nevertheless, by c a r e f u l l y evaluating the information presented, and
with a fundamental knowledge of noise sources and the e f f e c t of
sample-induced noises on an a n a l y t i c a l technique, the reader can
reach an i n t e l l i g e n t decision based on the information available.
Literature Cited
1. Epstein, M. S.; Winefordner, J . D. Prog. Analyt. Atomic
Spectrosc. 1984, 7, 67-137.
2. IUPAC Commission on Spectrochemical and Other Optical
Procedures for Analysis, Anal. Chem. 1976, 48, 2294-2296.
3. Zacha, K. E . ; Bratzel, M. P.; Winefordner, J . D.; Mansfield, J .
M. Anal. Chem. 1968, 40, 1733-1736.
4. Larkins, P. L. Spectrochim. Acta 1971, 26B, 477-489.
5. Capacho-Delgado, L.; Manning D C Spectrochim Acta 1966
22B. 1505-1513.
6. Kahn, H. L . ; Peterson
Newslett. 1968, 7, 35.
7. Gray, A. L. Spectrochim. Acta 1986, 41B, 151-167.
8. Turk, G. C. Anal. Chem. 1981, 53, 1187-1190.
9. Currie, L. A. Anal. Chem. 1968, 40, 586-593.
10. Long, G. L.; Winefordner, J . D. Anal. Chem. 1983, 55,
712A-718A.
11. Winefordner, J . D.; Ward, J . L. Anal. Lett. 1980, 13,
1293-1297.
12. Belchamber, R. M.; Horlick, G. Spectrochim. Acta 1982, 37B,
71-74.
13. Belchamber, R. M.; Horlick, G. Spectrochim. Acta 1981, 36B,
581.
14. Veillon, C. In Trace Analysis: Spectroscopic Methods for
Elements: Winefordner, J . D., Ed.; Wiley; New York, 1976;
Chapter 6, pp 164-166.
15. West, T. S.; Williams, X. K. Anal. Chim. Acta 1969, 45, 27.
16. Omenetto, N.; Winefordner, J . D. Prog. Analyt. Atom. Spectrosc.
1979, 2, 1-183.
17. Littlejohn, D.; Ottaway, J . M. Anal. Chim. Acta 1978, 98,
279-290.
18. IUPAC Commission on Spectrochemical and Other Optical
Procedures for Analysis, Appl. Spectrosc. 1977, 31, 348-364.
19. Boumans, P. W. J . M.; McKenna, R. J.; Bosveld, M. Spectrochim.
Acta 1981, 36B, 1031-1058.
20. Bower, N. W.; Ingle, J . D. Anal. Chem. 1977, 49, 574.
21. Bower, N. W.; Ingle, J . D. Anal. Chem. 1979, 51, 73.
22. Omenetto, N.; Winefordner, J . D. In Analytical Laser
Spectroscopy: Omenetto, N., Ed.; Wiley; New York, 1979;
23. O'Haver, T. C. In Trace Analysis: Spectroscopic Methods for
Elements: Winefordner, J . D., Ed.; Wiley; New York, 1976;
24. Weeks, S. J.; Haraguchi, H.; Winefordner, J . D. Anal. Chem.
1978, 50, 360-368.
25. Epstein, M. S.; Bayer, S.; Bradshaw, J.; Voigtman, E.;
Winefordner, J . D. Spectrochim. Acta 1980, 35B, 233-237.

26. Seltzer, M. D.; Hendrick, M. S.; Michel, R. G. Anal. Chem.

1985, 57, 1096-1100.
27. Epstein, M. S.; Travis, J.; Turk, G. C., in preparation.
RECEIVED December 24, 1986

Chapter 7
Noise and Detection Limits

in Signal-Integrating
Analytical Methods
1 2
H. C. Smit and H. Steigstra
1
Laboratory for Analytical Chemistry, University of Amsterdam, Nieuwe
Achtergracht 166, 1018 WV Amsterdam, Netherlands
2
Faculty of Medicine, Radboud Hospital, Nijmegen, Netherlands
The uncertainty i
is discussed, particularly the 0th moment (area) of a
peak determined via an integration procedure. An over-
view is given of the derivations of the error variance
due to integrated noise, both in the frequency domain
and in the time domain. As an example the uncertainty
in case of some typical kinds of noise is calculated,
using the derived expressions. The theory is extended
with the derivation of the optimum integration interval
on basis of known peak shapes and known noise characte-
r i s t i c s , assuming stationary noise without a determinis-
tic drift component. Finally, the influence of uncorrec-
ted linear drift on the integration variance is deter-
mined, while an expression for the variance after a
frequently applied drift correction is derived, using
correction intervals.
One o f the b a s i c problems i n a n a l y t i c a l c h e m i s t r y i s how t o c a l c u l a t e
the u n c e r t a i n t y i n t h e d e t e r m i n a t i o n o f t h e parameters o f a n o i s y
a n a l y t i c a l s i g n a l . Although t h i s u n c e r t a i n t y i s important, i t i s not
the o n l y f a c t o r i n f l u e n c i n g t h e d e t e c t i o n l i m i t . I t must be empha-
s i z e d t h a t e r r o r s and u n c e r t a i n t i e s o r i g i n a t i n g from sample p r e -
p r o c e s s i n g , sample i n t r o d u c t i o n , l a c k o f s t a n d a r d i z a t i o n o f t h e
measurement c o n d i t i o n s , e t c . , may be j u s t as i m p o r t a n t as n o i s e p e r -
t u r b i n g the s i g n a l . However, i t i s c e r t a i n l y u s e f u l t o c a l c u l a t e t h e
c o n t r i b u t i o n of that noise to the t o t a l u n c e r t a i n t y , determining the
d e t e c t i o n l i m i t , i n the a n a l y t i c a l r e s u l t .
I f o n l y one measurement ( d a t a ) p o i n t i s c o n s i d e r e d , then t h e
problem reduces t o a simple comparison o f the measured a m p l i t u d e w i t h
the s t a n d a r d d e v i a t i o n o f t h e n o i s e , determined by r e p e a t e d measure-
ments. O r d i n a r y s t a t i s t i c s can be a p p l i e d t o c a l c u l a t e t h e u n c e r t a i n -
t y . However, o f t e n dynamic s i g n a l s , l i k e peaks i n chromatography, a r e
produced and s i g n a l parameters l i k e t h e 0th moment (peak area) o r
h i g h e r moments a r e r e p r e s e n t a t i v e f o r t h e d e s i r e d a n a l y t i c a l i n f o r m a -
tion.
D e t e r m i n i n g these parameters always i n c l u d e s an i n t e g r a t i o n
0097-6156/88/0361 -0126$06.75/0

7. SMIT AND STEIGSTRA Noise and Detection Limits 127
p r o c e d u r e , where o f c o u r s e , t o g e t h e r w i t h t h e s i g n a l , the n o i s e i s
i n t e g r a t e d as w e l l . We might f o r m u l a t e the problem as f o l l o w s : What
i s the u n c e r t a i n t y i n the d e t e r m i n a t i o n o f the a n a l y t i c a l s i g n a l
parameters due t o the i n f l u e n c e o f the i n t e g r a t e d n o i s e ?
In t h i s paper we emphasize the d e t e r m i n a t i o n o f the peak a r e a s .
P a r t i c u l a r l y i n q u a n t i t a t i v e chromatography the u n c e r t a i n t y i n t h e
a r e a d e t e r m i n a t i o n i s d i r e c t l y r e l a t e d t o the d e t e c t i o n l i m i t . To
e l u c i d a t e the problem f o r m u l a t i o n , a (Gaussian) peak and t h e time
i n t e g r a l o f the peak i s shown i n F i g u r e 1. The r e l e v a n t i n f o r m a t i o n
i s the h e i g h t I o f the i n t e g r a l w i t h r e s p e c t t o b a s e l i n e , assuming a
c o n s t a n t ( f l a t ) n o i s e l e s s b a s e l i n e . I f n o i s e i s added and t h e peak i s
i n t e g r a t e d a g a i n , then the f i n a l v a l u e w i l l p r o b a b l y d i f f e r from t h e
t r u e v a l u e . R e p e a t i n g the same procedure w i t h a s i m i l a r peak w i t h
n o i s e w i t h the same s t a t i s t i c a l p r o p e r t i e s y i e l d s a number o f s t a t i s -
t i c a l l y d i s t r i b u t e d data p o i n t s . I f the n o i s e i s assumed t o be s t a -
t i o n a r y , i . e . i f the s t a t i s t i c a l p r o p e r t i e l i k d varianc
are n o t changing w i t h time
an e s t i m a t e o f the t r u e
v a r i a n c e a | determines the u n c e r t a i n t y i n t h e peak a r e a d e t e r m i n a -
tion.
The problem o f d e t e r m i n i n g the v a r i a n c e o f i n t e g r a t e d n o i s e i s
not r e s t r i c t e d t o peak parameter d e t e r m i n a t i o n . Measurement and
c a l c u l a t i o n o f the average i n t e n s i t y o f a s p e c t r o s c o p i c l i n e means
i n t e g r a t i n g t o o , however, the f i n a l r e s u l t i n c l u d i n g the s t a n d a r d
d e v i a t i o n o f the i n t e g r a t e d n o i s e has t o be d i v i d e d by the i n t e g r a -
t i o n time.
A l t o g e t h e r , t h i s b r i n g s us t o the d e s i r a b i l i t y t o d e r i v e an
e x p r e s s i o n f o r Oj_ o r a^, r e s p e c t i v e l y , c o n t a i n i n g a l l f a c t o r s i n f l u -
e n c i n g the e r r o r v a r i a n c e . Of c o u r s e , t h i s e x p r e s s i o n can be used t o
c a l c u l a t e the d e t e c t i o n l i m i t i n , f o r i n s t a n c e , chromatography as f a r
as determined by the b a s e l i n e n o i s e . However, i t i s a l s o u s a b l e t o
make an optimum c h o i c e o f parameters and c o n d i t i o n s . B e s i d e s , some
r u l e s o f thumb can be g i v e n , u s a b l e i n d a i l y p r a c t i c e .
One has t o keep i n mind t h a t such a d e r i v a t i o n always i m p l i e s
some assumptions c o n c e r n i n g the s t a t i o n a r i t y o f the a n a l y t i c a l system
and p a r t i c u l a r l y the s t a t i o n a r i t y o f the n o i s e . I n g e n e r a l , s t a t i o -
n a r i t y and the absence o f a d e t e r m i n i s t i c d r i f t i n g b a s e l i n e i s
assumed, a l t h o u g h some d e r i v e d e x p r e s s i o n s i n the g e n e r a l form a r e
v a l i d f o r n o n - s t a t i o n a r y n o i s e . However, t h e d e r i v e d t h e o r y can be
used as a b a s i s f o r t h e c a l c u l a t i o n o f t h e r e m a i n i n g u n c e r t a i n t y i n
the case of a c o r r e c t i o n procedure f o r d e t e r m i n i s t i c ( f o r i n s t a n c e
linear) baseline d r i f t .
B a s i c Theory
The d e r i v a t i o n o f the e r r o r v a r i a n c e r e q u i r e s some t h e o r y from d i f f e -

r e n t f i e l d s . F o r the convenience o f t h e reader a v e r y s h o r t o v e r v i e w
w i l l be g i v e n , i n c l u d i n g some b a s i c p r i n c i p l e s and d e f i n i t i o n s o f the
r e q u i r e d t h e o r y . Another r e a s o n t o g i v e some textbook t h e o r y i s t h a t
the d e f i n i t i o n o f s e v e r a l q u a n t i t i e s can d i f f e r ; l i t e r a t u r e i s n o t
very c o n s i s t e n t i n that respect. A d e t a i l e d d e s c r i p t i o n of s i g n a l
t h e o r y , system t h e o r y , s t o c h a s t i c p r o c e s s e s and o f course mathematics
can be found i n s e v e r a l textbooks ( 1 - 5 ) .
I t i s n e c e s s a r y to s o l v e t h e problem o f d e r i v i n g b o t h i n the

Figure 1. Integrated peak and noise.

time domain and i n the f r e q u e n c y domain by F o u r i e r t r a n s f o r m i n g the

time s i g n a l s o r f u n c t i o n s d e r i v e d from t h e time s i g n a l s . The F o u r i e r
t r a n s f o r m (FT) o f a f u n c t i o n o f the time f ( t ) i s d e f i n e d i n t h e u s u a l
way:
+ 00
F(jco) = f f(t) e" J U 3 t
dt (1)
CO
2
j = -1
We w i s h t o c o n s i d e r random v a r i a b l e s i n a c o n t i n u o u s domain. T h i s can

be done by u s i n g d i f f e r e n t d e s c r i p t i v e f u n c t i o n s . F o r o u r d e r i v a t i o n s
we need the concepts o f p r o b a b i l i t y d e n s i t y f u n c t i o n (PDF), a u t o -
c o r r e l a t i o n f u n c t i o n (ACF) and power s p e c t r a l d e n s i t y (PSD). More-
o v e r , the f o l l o w i n g system f u n c t i o n s a r e used: t h e w e i g h t i n g f u n c t i o n
h ( t ) and t h e complex f r e q u e n c
The w e l l known p r o b a b i l i t
the l i m i t i n g v a l u e o f t h e p r o b a b i l i t y t h a t an a m p l i t u d e o f n o i s e n ( t )
l i e s i n an i n t e r v a l around a c e r t a i n v a l u e , d i v i d e d by the w i d t h o f
t h a t i n t e r v a l . The shape, which i s o f t e n G a u s s i a n , and t h e w i d t h o f
the PDF, e x p r e s s e d i n the s t a n d a r d d e v i a t i o n a , a r e used f o r s t a t i s -
n
t i c a l c a l c u l a t i o n s o f d e t e c t i o n l i m i t e t c . A random s i g n a l , o r i n
g e n e r a l a f a m i l y o f f u n c t i o n s o f time (random p r o c e s s ) o f w h i c h t h e
v a l u e s v a r y randomly even i f i t i s s t a t i o n a r y , i s n o t u n i q u e l y
s p e c i f i e d by a PDF, as i s demonstrated i n F i g u r e 2. Both random
s i g n a l s have the same PDF, b u t they a r e o b v i o u s l y d i f f e r e n t .
An i m p o r t a n t q u a n t i t y , summarizing much i n f o r m a t i o n about a
random p r o c e s s , i s the ACF. To i l l u s t r a t e the concept o f t h e ACF,
F i g u r e 3 shows a f a m i l y o f s t o c h a s t i c s i g n a l s ( s i g n a l s e v o l v i n g i n
time a c c o r d i n g t o p r o b a b i l i t y l a w s ) , a random ( s t o c h a s t i c ) p r o c e s s
o r an ensemble. An example o f an ensemble i s a s e t o f p o s s i b l e n o i s e
r e c o r d s from a chromatographic d e t e c t o r , each r e c o r d e d d u r i n g a
c e r t a i n time i n t e r v a l . Now we have t o d i s t i n g u i s h ensemble s t a t i s t i c s
and time s t a t i s t i c s . F o r i n s t a n c e , t h e mean v a l u e a t t h e time t i
(ensemble s t a t i s t i c s ) i s d e f i n e d :
N
1
u( t l ) = lim 1 E n, ( t l ) (2)
k r e f e r s t o s i g n a l k.
The (ensemble) ACF i s d e f i n e d :
N
1
R (ti,t!+T)
n = R(ti,t ) = lim - I n (t )n (ti+T)
2 k 1 k (3)
0 0
N k= 1
b e i n g the average p r o d u c t o f the v a l u e s o f t h e s t o c h a s t i c p r o c e s s a t

time t i and t 2 . I f t h e r e i s no r e l a t i o n o r b e t t e r c o r r e l a t i o n between
the v a l u e s a t time t i and t , then the average w i l l tend t o t h e p r o -
2
duct o f the mean v a l u e a t t i and t . I n case o f f a s t f l u c t u a t i n g

2
n o i s e , no c o r r e l a t i o n w i l l e x i s t even a f t e r a r e l a t i v e l y s h o r t t i m e ,
where s l o w l y f l u c t u a t i n g n o i s e s t i l l shows an average r e l a t i o n
between t h e a m p l i t u d e s .

Figure 2. Fast and slowly fluctuating noise with similar PDF.

Figure 3. Ensemble of noise records.

A stochastic process x^(t) i s stationary i f the two processes

x(t) and x ( t + e ) have the same s t a t i s t i c s f o r any . In other words,
the values of t i and t do not influence the ensemble s t a t i s t i c s ,
2
only the difference t i - t = T i s important, R ( t i , t ) can be replaced 2 2
by R ( t i ~ t ) . 2^(t) c a n D e
^ noise records n^ or a set of a s e t
other random variables. If the expected value E[x(t)] = ]i - constant,

and E [ x ( t + T ) x ( t ) ] = R ( T ) i s only dependent on T and does not vary
with the time ( t i ~ t i s replaced by T ) , then the process i s weakly
2
stationary.
Ergodicity means: a l l s t a t i s t i c s can be determined from a single
function x ^ ( t ) :
u (k) = lim i f x, (t)d (4)
and
R X X ( T , k ) = lim 1 j x^t) x^t + 'Odt (5)
If T = 0, Equation 5 reduces to:

T
2 2
R(0) = l i m i [ x (t)dt = E[x (t)] (6)
being the mean square value of x ( t ) , determining the average power

and thus the energy of the signal.
Of course, fast and slowly fluctuating noise can also be d i s -
tinguished i n the frequency domain. However, noise usually goes on
i n d e f i n i t e l y i n time and, actually, i t s energy i s unbounded; the
Fourier transform does not e x i s t . Nevertheless, FT techniques can be
applied i f the average power i s bounded. The introduction of the
power spectral density function (PSD), not suitable f o r a simple
representation of the (stochastic) signal, allows the introduction of
expressions concerning signal energy. The PSD gives the average power
per unit of frequency. Of course, the ACF and the PSD are not inde-
pendent; s t r i c t l y speaking, both are representing the same properties
of the signal (energy, fast or slow f l u c t u a t i o n ) . The formal d e f i n i -
tion of the PSD i s :
S(u>) = FT {R X X (T)| (7)
Because R ( T ) i s real and even for real processes, S(OJ) i s also r e a l

and even, and:
+ 00
S(OJ) = [
J R
xx
(T) COS CUT dx = 2
Jf R
xx
(T) COS COT dT (8)

F i n a l l y , the physical r e a l i z a b l e one-sided PSD function G(to) i s

defined as:
oo
G(o)) = 4 j R X X (T) cos OJT dr (9)
The energy of the signal can be calculated from G(co):
oo
E[x (t)] = J
2
G(u>) do) = R^CO) (10)
0
2 2
E [ x ( t ) ] i s the mean square value i(/ of the signal x ( t ) . I f the mean
of x ( t ) i s not zero, f o instanc i f x(t) i nois with Direct
Current (DC) component
t o t a l energy. However,
variations of the signal and not i n the mean value which can be
estimated and corrected. Therefore, i n the following we assume a mean
2
value of zero, i n which case E [ x ( t ) ] becomes the variance a .
2
Some d e f i n i t i o n s from l i n e a r system theory are required. The

weighting function h(t) i s the response (output) of a l i n e a r system
applied to an impulse s i g n a l , t h e o r e t i c a l l y a Dirac-delta function
6 ( t ) , or more precisely, a 6-distribution with the properties:
+ oo
6(t) dt
(ID
6(t) = 0 (t + 0)
The FT of the impulse response i s the complex frequency response

H(joj). Suppose x(t) with a PSD = G (OJ) i s the input signal of a X
l i n e a r system with a complex frequency response H(jio) and suppose

y(t) i s the r e s u l t i n g output. Then i t can be proved (J_) that the PSD
of y ( t ) i s given by:
2
G (u>) = |H(jco)| G (o>) (12)
y x
Variance of Integrated Noise
The system functions, mentioned i n the previous paragraph, can be

determined for an integrator. The response of an integrator applied
to an impulse 6(t) i s the value 1 ( t > 0 ) , as follows from the d e f i -
n i t i o n of the 6-distribution (Equation 11). The FT can e a s i l y be
calculated, r e s u l t i n g i n :
-joot J _
(13)
H(jO)) = FT | h ( t ) | = J h(t) e j U ) t
dt = J dt =

A c c o r d i n g t o E q u a t i o n 12 we c a n c a l c u l a t e t h e PSD o f t h e o u t p u t ,
assuming a s t o c h a s t i c i n p u t s i g n a l w i t h known PSD:
G (0)) = |H(ja))| G (co) = - L (u>) 2

G (14)
y x ^2 x
At f i r s t s i g h t , i t appears p o s s i b l e t o c a l c u l a t e , w i t h o u t any p r o -
blem, t h e v a r i a n c e o f t h i s o u t p u t s i g n a l by i n t e g r a t i n g the c a l c u l a -
ted output PSD over the 0)-range from zero t o (see E q u a t i o n 10): 0 0
?
k i
V = f - L G (co)dco (15)
J u) 2 x
However, t h i s i s n o t c o r r e c t , as a l r e a d y i s shown i n p r e v i o u s papers

( 6 , 7 ) . A s i g n a l i s neve
a l i m i t e d time i n t e r v a l
In r e a l i t y , t h e impulse response h ( t ) o f an i n t e g r a t o r i s n o t 1, b u t
i s g i v e n by:
h(t) = 1 (0<h(t)<T) ( 1 6 )
= 0 (else)
T i s the i n t e g r a t i o n i n t e r v a l .
F o u r i e r t r a n s f o r m i n g E q u a t i o n 16 g i v e s a d i f f e r e n t e x p r e s s i o n f o r
H(jco):
H(JOJ) = h(t) e J
dt = : (17)
J
CO
The PSD of t h e output s i g n a l i s now:
G(OJ) = lH(jo))l G (o)) = 2

Y
s i n
^ / T 2
G (co) (18)
y X 2 X
(co/2)
The error variance i s :
- r sin (ojT/2) _ ,
2
, , , sin 0)T/2
2
, coT
V T - f S l
<*T/2) G (co) dco - 2T
n N O T
f G (a)) d ^ (19)
((0/2) 2
^ J A
(u)T/2) 2 2
One has t o keep i n mind t h a t a s i g n a l w i t h a PSD g i v e n by E q u a t i o n 18

i s n o t p r e s e n t a t t h e output o f a s i n g l e i n t e g r a t o r . I t i s n e c e s s a r y
to c o n s i d e r a s e t o f p o s s i b l e outcomes o f k s i m i l a r i n t e g r a t i o n
p r o c e d u r e s , i n o t h e r words, t o use t h e a l r e a d y mentioned concept o f
an ensemble. The k i n E q u a t i o n 19 denotes an ensemble r e p r e s e n t a t i o n .
As a s i m p l e a p p l i c a t i o n o f E q u a t i o n 19, b a s e l i n e n o i s e w i t h a r e c -
t a n g u l a r PSD, i . e . w h i t e n o i s e w i t h energy d i s t r i b u t e d u n i f o r m l y over
a l l f r e q u e n c i e s b a n d l i m i t e d by an i d e a l low pass f i l t e r w i t h r e c t a n g -
u l a r passband, w i l l be t r e a t e d . I n f o r m u l a :

6(a)) = K (0<o)<a) )
(20)
G(OJ) = 0 (a>>u> ) 0
K = constant.
This k i n d of noise i s not very r e a l i s t i c , a true i d e a l f i l t e r i s
" r e a l - t i m e " i m p o s s i b l e . However, such a spectrum can be approximated
w i t h a h i g h e r o r d e r f i l t e r w i t h sharp c u t - o f f c h a r a c t e r i s t i c s .
The v a r i a n c e o f t h e n o n - i n t e g r a t e d b a s e l i n e n o i s e can be c a l c u l a t e d ,
u s i n g E q u a t i o n 10:
^0
Q = G a ) d a ) =
n j' ( ) /G(oj)da) (21)
0 0
resulting i n :
2
a
K = (22)
%
Substituting Equation 22 and Equation 20 into Equation 19 we obtain:
u T/2
0
V . 0 2 . 2T r sin (0)T/2)
2
( 2 3 )
1 n 2
% J ( T/2)
W
g i v i n g a q u a n t i t a t i v e e x p r e s s i o n f o r the v a r i a n c e o f i n t e g r a t e d
noise.
E q u a t i o n 23 does n o t l o o k v e r y a t t r a c t i v e f o r r o u t i n e u s e , b u t o f t e n
i t can be s i m p l i f i e d . F o r example, 00QT/2 has a r e l a t i v e l y l a r g e v a l u e
i n chromatography, as w i l l be proved. L e t us assume chromatographic
peaks w i t h a G a u s s i a n peak shape and s t a n d a r d d e v i a t i o n Cfp, d e t e r -
m i n i n g t h e peak w i d t h . The minimum i n t e g r a t i o n i n t e r v a l w i t h an
a c c e p t a b l e s y s t e m a t i c e r r o r (< i n the area determination i s
about 7 Gp. The f r e q u e n c y spectrum o f t h e peak can be determined by
F o u r i e r t r a n s f o r m i n g t h e G a u s s i a n peak f u n c t i o n . The r e s u l t i s a l s o a
G a u s s i a n f u n c t i o n i n t h e f r e q u e n c y domain, however, w i t h a s t a n d a r d
deviation = l / a . To p r e v e n t u n a c c e p t a b l e peak d i s t o r t i o n , t h e
p
minimum c u t - o f f f r e q u e n c y OJQ o f a low pass f i l t e r has t o be about

3^.0^; h i g h e r f r e q u e n c i e s can be n e g l e c t e d i n case o f a G a u s s i a n PSD.
Hence:
v J ' y 7 g
p ,
M2 (24)
In p r a c t i c e , 0JQ i s determined by t h e peak w i t h t h e smallest peak w i d t h .

A l l t h e o t h e r peaks r e q u i r e a l a r g e r i n t e g r a t i o n time and 0)QT/2 > 12.
The c o n t r i b u t i o n t o t h e i n t e g r a l r e s u l t i n g from t h e i n t e r v a l between
T
U)0 /2 ( > 1 2 ) and i n f i n i t e i s n e g l i g i b l e . The v a l u e o f t h e i n t e g r a l
00
w i t h i n t e g r a t i o n l i m i t s 0 and i s TT/2. The f i n a l r e s u l t i s :

k 2 2
a - a & (25)
I n u) Q
In this p a r t i c u l a r case the variance of the integrated noise i s pro-

portional to the integration i n t e r v a l and to the variance a of the n
A R E
o r i g i n a l baseline noise. We note that a and COQ 22 independent, n
reducing COQ means reducing o . Noise with a strong 1/f character i s

n
much more r e a l i s t i c (7), p a r t i c u l a r l y i n chromatography. The PSD of

1/f (or 1 /to) noise i s proportional to 1/to. Because of the s i n g u l a r i t y
in co=0, a s l i g h t l y modified model i s more r e a l i s t i c . Such a PSD
might be:
G(u)) = K/oj 0 < oo < u>

(26)
G(oj) = K/00 0) < co
co^ i s a fixed (low) frequency
Substitution i n Equation 19 gives:
k2 a m 2 T JL f 2 2
sin (o)T/2) d a ) T / 2 + 2 k t 2 r" s i n ( 0 T / 2 )
2
t d ( ( 0 T / 2 ) ( 2 y )
J 2 3
Q (a)T/2) J (a)T/2)
For low values of u) the f i r s t term i s approximately KT , the i n t e -

gral i n the second term has to be calculated numerically. An important
conclusion i s that i n case of 1/f ( f l i c k e r ) noise the variance o\ i s
2
proportional to T .
A treatment i n the frequency domain i s not always optimal. For
instance, calculations with non-stationary stochastic processes are
d i f f i c u l t . Moreover, the PSD i s mostly determined by Fourier trans-
forming the ACF, therefore an expression using d i r e c t l y the ACF
avoids Fourier transforming. The derivation happens to be not d i f f i -
cult. Assuming a random signal n(t) with mean value y, we write the
random v a r i a b l e :
b b
I - J n(t)dt E[n(t)] = \i E[I] = f y(t)dt (28)
Now we write:
b b
I 2
- J n(ti)dti J n(t )dt 2 2 (29)
a a
using two dummy (time) variables t i and t . 2
The integration l i m i t s are independent and the iterated integral

can be written as a double i n t e g r a l :

b b
I 2
= J Jn(t )n(t )dt dt
1 2 1 2 (30)
a a
Taking the expected value and interchanging the expected value proce-
dure and the integration gives:
b b
E[I ] 2
= f |E[n(t )n(t )]dt dt 1 2 1 2 (31)
a a
b b
2
R(ti,t )dt!dt 2 2 = a (32)
a a
More or less naturally th
usable expression. I f , , y ergodicity
assumed, then we get:
T T
2
E[I ] 2
- f 2
j R(t -t )dt dt
1 2 1 2 (33)
T T
~ 2 "2
T T
assuming an integration i n t e r v a l from - to .
Equation 33 can be simplified to:
T
E[I ] 2
= 2j (T-T)R(T)dT (34)
0
T = ti - t . 2
The proof, being purely mathematical, i s omitted (j6).

As a f i n a l r e s u l t we have two expressions (Equations 19 and 34)
with certain r e s t r i c t i o n s quite usable i n practice. One can prove
that the expressions e s s e n t i a l l y are the same and they can be derived
from one another.
Figures 4 and 5 show some t y p i c a l types of noise: f i r s t order
noise, i . e . white noise bandlimited by a simple f i r s t order f i l t e r
with time constant T i , and noise with a strong 1/f component, o r i g i -
nating from an Inductively Coupled Plasma - Atomic Emission Spectro-
graph (ICP-AES).
The ACF of f i r s t order noise i s :
-Kl
R ( T ) = a* exp - - (35)
n Ti
Substitution i n Equation 34 r e s u l t s i n :

Figure 4. ACF, PSD and record of f i r s t order noise.

noise I C P
0.0
( )
eg
o
0.00 4.00 8.00 12.00 16.0

1
T (XIO )
40.00 r
32.00 PSD noise I C P
24.00
I
16.00
><
8.00
X
a! 0.00
-8.00
0.00 1.60 320 4.80 6.40
1
Freq. U I0 )
F i g u r e 5. ACF, PSD and r e c o r d o f n o i s e o f an I n d u c t i v e l y Coupled

Plasma - Atomic E m i s s i o n S p e c t r o g r a p h .

o\ = a* j^2TT 1+ 2Tf jexp (-^-) - l J J (36)
In chromatography T T i , and the expression can be simplified to:
2
a ? . a . 2TTi (37)
l n
2
Again, the variance i s proportional to T, a and T i are not indepen-
dent. Figure 6 shows the r e s u l t of the computer c a l c u l a t i o n of the
error variance as a function of the integration time. The o r i g i n of
the noise i s a Flame Ionisation Detector (FID). The ACF i s estimated
from a limited number of baseline noise data points, r e s u l t i n g i n a
confidence i n t e r v a l derived with the B a r t l e t t formula (4,8).
An i n t e r e s t i n g question i s : Is f i l t e r i n g p r i o r to integration
useful, p a r t i c u l a r l y i
decreasing cut-off frequenc
of the baseline noise an ,
decreasing cut-off frequency r e s u l t s i n a distorted peak; the peak
width i s increasing and implying that an increasing integration time
i s needed to avoid systematic errors; as i s shown, t h i s i s not
favorable. A study carried out with f i r s t order and second order
f i l t e r s has shown that f i l t e r i n g p r i o r to integration i s not advisa-
ble, as the second effect dominates (6).
Optimum Integration Limits
Another interesting question i s : Is i t possible to determine optimal

peak integration i n t e r v a l s on the basis of known or even unknown peak
shapes and known noise characteristics? And i f an optimum integration
i n t e r v a l can be estimated, what i s the error variance?
As i s extensively shown i n a previous paper (9), i t happens to
be possible i n some cases to determine optimum integration l i m i t s .
As an example l e t us consider a symmetric peak with, f o r instance,
a Gaussian peak shape with known peak maximum. Decreasing the i n t e -
gration i n t e r v a l means decreasing the random error i n the peak area
estimate, as i s shown. But the systematic error i s increasing; the
peak i s not completely integrated and the r e s u l t i n g area w i l l be
biased.
Figure 7 shows a signal x ( t ) composed of a peak s(t) (dashed
line) and noise n ( t ) . 1^ i s the true peak area and !, i s t h e peak
area estimate, taken as the area of the noisy peak within the i n t e -
gration i n t e r v a l divided by the normalized integrated peak f r a c t i o n
I ( T ) = I (T)/Ioo> being the f r a c t i o n of the peak area i n the
n o r m S
i n t e r v a l T divided by the true peak area. This i s equal to the shaded

part of the small peak with unit area.
We want to minimize P 2 ( T ) , i . e . the expected value of the
squared difference between the true area and the estimate:
( T )
I
y (T) = E | I - I 1
2 = = error variance (38)
I 00 00 I
norm

Figure 7. Noisy peak and normalized peak with optimum integration

interval.

() has t o be m i n i m i z e d w i t h r e s p e c t t o T. The n o r m a l i z e d i n t e g r a
2
ted peak a r e a i n case o f a known peak shape i s known, and we have

d e r i v e d an e x p r e s s i o n f o r t h e i n t e g r a t e d n o i s e v a r i a n c e . As u s u a l , we
can determine the d e r i v a t i v e , s e t t i n g i t e q u a l t o z e r o and t h e
d e s i r e d r e s u l t can be c a l c u l a t e d . T h i s p r o c e d u r e l e a d s t o :
da* (T)
( T )
I'(T) - - 2* (T) d I
' (39)
dT dT
In case of symmetric peaks:

u+^T
R ( T ) =
K S " T~ f
= s ( t ) d ( t ) ( 4 0 )
u-^T
An e x p r e s s i o n f o r the v a r i a n c e i s known ( E q u a t i o n 34):

* () = 2 ( - ) R () (41)
I
Hence, u s i n g E q u a t i o n s 40 and 41 E q u a t i o n 39 can be w r i t t e n a s :

u+^T s ( t ) d t . R () = 2 f ( T - ) R n T n
() . [ s ( u + i T ) ] (42)
J J n n
J
u-^T 0 0
E v a l u a t i o n g e n e r a l l y l e a d s t o a minimum f o r some v a l u e o f T. S i m i l a r
e q u a t i o n s can be d e r i v e d f o r asymmetric peaks ( 9 ) .
The f i n a l r e s u l t i n case o f a G a u s s i a n peak w i t h f i r s t o r d e r
noise i s :
e r f [ - J L - U V T . X e x p i - ^ l ) (43)
= s t a n d a r d d e v i a t i o n o f t h e peak.
This r e l a t i o n i s s a t i s f i e d i f :
2.8 (44)
opt
The c o r r e s p o n d i n g error variance i s :
2
5.6
(
2 J - ? - J L ~ 8 o 2 T (45)
2 n X P
p t
erf (0.99)
= time c o n s t a n t of the f i r s t o r d e r n o i s e ,


A remarkable r e s u l t i s t h a t i n t h i s p a r t i c u l a r case the optimum

i n t e g r a t i o n i n t e r v a l i s independent of the time c o n s t a n t and of the
v a r i a n c e of the n o i s e . Of c o u r s e , the r e s u l t i n g e s t i m a t i o n e r r o r
depends on b o t h parameters. The t h e o r y can be extended t o skewed
peaks w i t h known or unknown shape, o t h e r k i n d s of n o i s e , e t c .
A d e t a i l e d treatment i s g i v e n i n ( 9 ) .
Integration Variance after Baseline Correction
The e x p r e s s i o n s d e r i v e d so f a r are o n l y v a l i d i f the n o i s e i s assumed

to be s t a t i o n a r y . However, i t i s u n f o r t u n a t e t h a t t h i s i s not always
the case. P a r t i c u l a r l y i n a technique l i k e chromatography, a non-
s t a t i o n a r y b a s e l i n e d r i f t i s o f t e n p r e s e n t due t o , f o r i n s t a n c e ,
s t r i p p i n g of the column, c o n t a m i n a t i o n of d e t e c t o r s , e t c . The non-
s t a t i o n a r y d r i f t , not t o be confused w i t h s t a t i o n a r y low frequency
noise with properties define i p r o b a l i s t i
c o n s i d e r e d as a d e t e r m i n i s t i
Baseline d r i f t correctio indispensabl par goo
chromatographic d a t a p r o c e s s i n g p r o c e d u r e . The f o l l o w i n g q u e s t i o n s
have t o be answered:
- What i s the i n f l u e n c e of u n c o r r e c t e d d r i f t on the e s t i m a t e d ACF,
which i s used to c a l c u l a t e s t a t i s t i c a l q u a n t i t i e s l i k e the i n t e g r a
t i o n error variance?
- What i s the i n t e g r a t i o n v a r i a n c e a f t e r d r i f t c o r r e c t i o n ?
A complete treatment i s beyond the scope of t h i s paper, but an i n t r o
d u c t i o n w i t h a s i m p l i f i e d , but p r a c t i c a l l y r e l e v a n t example, w i l l be
g i v e n here. The s i m p l e s t case i s a l i n e a r b a s e l i n e d r i f t and the
r e s u l t i n g model of the n o i s y d r i f t i n g b a s e l i n e i s :
x(t) = n ( t ) + a + bt (46)
where a and b are c o n s t a n t s . m

A f i n i t e measurement time from - to i s chosen, w h i c h i s not
important f o r the d e r i v a t i o n .
The e s t i m a t i o n of the ACF i s now d e f i n e d as:

(47)

2
E v a l u a t i o n of t h i s i n t e g r a l g i v e s :
true ACF systematic e r r o r random e r r o r
R* () = R () + a 2
+ (48)
xx xx
and Q are s t o c h a s t i c f u n c t i o n s of , b o t h w i t h an expected v a l u e of

zero:

() n(t)dt + n(t + x)dt (49)

(-)

T -
1 r2 1 ,2" 2
Q(T) = [ L
n(t) t dt +
n(t)dt + ^ J n(t + T ) t d t

' 2
-1 4 (50)
However, assuming , t h e i r v a r i a n c e can be d e t e r m i n e d as t h e

r a t h e r simple expression:
-
8 J
2
E[P (T)]! (--t
and
[Q ( )] (2 ( - ) / 3 ) J " ( t ) dt
2
(52)
As an example, t h e e q u a t i o n s f o r f i r s t o r d e r n o i s e w i l l be g i v e n :
2
E [ P ] = 8 /(-) 2
(53)
2
E[Q = 2 2
(-)/3 (54)

A c l o s e l o o k a t E q u a t i o n 48 l e a d s t o t h e f o l l o w i n g c o n c l u s i o n c o n
c e r n i n g t h e e f f e c t o f u n c o r r e c t e d l i n e a r d r i f t . The e s t i m a t e d ACF
2 2
c o n t a i n s two s y s t e m a t i c components, each p r o p o r t i o n a l t o a and b
r e s p e c t i v e l y , and two s t o c h a s t i c components, p r o p o r t i o n a l t o a and b.
A f i n a l c o n c l u s i o n can be d e r i v e d from t h e f o r m u l a e : a c o n s i d e r a b l e
e r r o r i n the e s t i m a t i o n o f t h e ACF and d e r i v e d q u a n t i t i e s can be
expected i f b a s e l i n e d r i f t i s n o t c o r r e c t e d . T h i s l e a d s us t o t h e
r e m a i n i n g q u e s t i o n , the d e t e r m i n a t i o n o f t h e i n t e g r a t i o n v a r i a n c e
after baseline d r i f t correction.
Many c o r r e c t i o n p r o c e d u r e s a r e known: l i n e a r and e x p o n e n t i a l f i t t i n g ,
p o l y n o m i a l a p p r o x i m a t i o n ( b o t h o r t h o g o n a l and n o n - o r t h o g o n a l ) , e t c .
I n t h i s paper the d e s c r i p t i o n w i l l be r e s t r i c t e d t o the v e r y o f t e n
used l i n e a r e x t r a p o l a t i o n , a g a i n assuming a l i n e a r b a s e l i n e d r i f t .
L e t us c o n s i d e r a n o i s y peak w i t h a l i n e a r d r i f t i n g b a s e l i n e . The
u s u a l p r o c e d u r e i s as f o l l o w s ( F i g u r e 8). Two time i n t e r v a l s w i t h a
time d u r a t i o n T a r e s e l e c t e d on b o t h s i d e s o f t h e peak. Now each
c
i n t e r v a l i s f i t t e d w i t h a s t r a i g h t l i n e . To s i m p l i f y t h e e q u a t i o n s
the f o l l o w i n g i n t e g r a l s a r e d e f i n e d :


T 2 T 3
1 1 = Jx(t)dt I 2 = J x(t)dt I 3 = J x(t)dt (55)
Ti T 2 T 3
The l e n g t h o f t h e i n t e g r a t i o n i n t e r v a l s a r e T , T^ and T , r e s p e c t i c c
v e l y . The b a s e l i n e d r i f t c o r r e c t e d i n t e g r a l i s :
(I1 + I3)
I = I 2 - - i (56)
2T
c
The v a r i a n c e of t h e c o r r e c t e d i n t e g r a l can be c a l c u l a t e d :
2 2
di.i 2 + . )
2 3 /( 1 + ) \ ~|
3
2
.\ .
( 2TJ -([. ]+[ . ]) ^
2 2 3 +
?
+ [ 1 ]
3 (57)
The f i r s t t h r e e e x p e c t e d v a l u e s can be d e t e r m i n e d by a l r e a d y d e s c r i
bed o r d i n a r y "|" c a l c u l a t i o n s . The o t h e r t h r e e a r e a c t u a l l y c r o s s
terms.
E v a l u a t i o n leads to a r a t h e r complicated formula, but nevertheless i t
i s g e n e r a l l y u s a b l e f o r a l l k i n d s of n o i s e w i t h known ACF. As an
example t h e r e s u l t f o r f i r s t o r d e r n o i s e w i l l be g i v e n :
2 2
= 2T.X -2x n
2
(l - e x p ( - T . / x ) ) - n " p(- W ) 2
.
I
e x p ( - ./ ) 2 (l - e x p ( - T ^ )
+ -T / r j l -L exp(- T./tJ).
(l - e x p ( - T / x ) ) c n (58)
E q u a t i o n 58 i s g r a p h i c a l l y d i s p l a y e d i n F i g u r e 9, showing -j- as a
f u n c t i o n of T^ w i t h t h e c o r r e c t i o n i n t e r v a l as a parameter.
A remarkable r e s u l t i s that the curves are c r o s s i n g .
A c a r e f u l i n s p e c t i o n of E q u a t i o n 58 and F i g u r e 9 l e a d s t o t h e f o l l o w
i n g statement: I f a s i g n a l w i t h l i n e a r d r i f t i n g b a s e l i n e and f i r s t
o r d e r b a s e l i n e n o i s e i s i n t e g r a t e d , then the optimum b a s e l i n e c o r r e c
t i o n i n t e r v a l i s i n f i n i t e i f t h e i n t e g r a t i o n time i s g r e a t e r than
f o u r times the time c o n s t a n t o f t h e n o i s e ; o t h e r w i s e , the optimum
c o r r e c t i o n i n t e r v a l i s z e r o . I n t h e l a s t case the use o f two c o r r e c
t i o n p o i n t s on b o t h s i d e s o f a peak i s s u f f i c i e n t .

.

F i g u r e 9. Standard d e v i a t i o n o f t h e i n t e g r a t e d n o i s e a f t e r d r i f t
c o r r e c t i o n v e r s u s the i n t e g r a t i o n t i m e , w i t h t h e c o r r e c t i o n
i n t e r v a l w i d t h as a parameter.

Library
In Detection in1155 16thChemistry;
Analytical S t , N.W.Currie, L.;
Washington,
ACS Symposium Series; American D.C
Chemical 20036
Society: Washington, DC, 1987.
Discussion and Conclusions
In case of a stable and stationary chromatographic system, the

derived general theory i s certainly usable for calculating quantita-
t i v e l y uncertainties and detection l i m i t s i n signal integrating
methods l i k e chromatography. However, an extensive analysis of the
detector noise i s required and the use of a computer with a data
acquisition system and special software i s inevitable.
If a d r i f t i n g baseline i s present and possibly corrected, the
formulae become rather complicated and are not d i r e c t l y usable i n
d a i l y practice. An extension to other kinds of noise, i . e . the more
r e a l i s t i c 1/f or f l i c k e r noise leads to even more complicated f o r -
mulae. Nevertheless, i t i s possible to determine quantitatively
detection l i m i t s i n case of the application of some s p e c i f i c baseline
d r i f t correction procedure, i f the measurement conditions are well
defined and stable and
from an a p r i o r i analysis
effects influencing th g y
i s obtained.
Literature Cited
1. Bendat, J. S.; Piersol, A. G. Measurement and Analysis of
Random Data; Wiley: New York, 1966.
2. Box, G. E. P.; Jenkins, G. M. Time Series Analysis; Holden-Day:
San Francisco, 1976.
3. Papoulis, A. The Fourier Integral and its Applications;
McGraw-Hill: New York, 1962.
4. Jenkins, G. M.; Watts, D. G. Spectral Analysis and its Appli-
cations; Holden-Day: San Francisco, 1969.
5. Beauchamp, K.; Yuen, C. Digital Methods for Signal Analysis;
Allen & Unwin: London, 1979.
6. Smit, H. C.; Walg, H. L. Chromatographia 1975, 8, 311.
7. Smit, H. C.; Walg, H. L. Chromatographia 1976, 9, 483.
8. Duursma, R. P. J . ; Smit, H. C. Anal. Chim. Acta 1981, 133, 67.
9. Laeven, J. M.; Smit, H. C. Anal. Chim. Acta 1985, 176, 77.
RECEIVED January 21, 1987

Chapter 8
Establishing Clinical Detection Limits

of Laboratory Tests
Mark H. Zweig
Clinical Pathology Department, Clinical Center, National Institutes of

Health, Bethesda, MD 20892
Fundamental clinical laboratory test performance can

be described in terms of accuracy, or the ability to
correctly classify subjects into clinically relevant
subgroups. Receiver operating characteristic (ROC)
curves demonstrate the limits of a given test to
detect the alternative states of interest over the
complete spectrum of operating conditions, providing
a comprehensive and pure index of accuracy.
Obtaining valid data for ROC analysis requires
attention to the following important steps: (1)
define carefully the specific clinical question to be
addressed; (2) choose subjects who are representative
of the population to which the test is ultimately to
be applied; (3) perform all tests being evaluated on
all subjects; (4) determine the "true" diagnosis by
rigorous and complete means independent of the
test(s) being studied; and (5) evaluate and compare
test performance at all decision levels using ROC
curves.
Swets and Pickett ( 1 ) divide t e s t performance into a discrimination

or accuracy aspect and a decision or e f f i c a c y aspect. Accuracy,
on the one hand, r e f e r s to the a b i l i t y of the test to c l a s s i f y , to
c o r r e c t l y discriminate between alternative c l i n i c a l states of the
subjects under study ( i . e . , signals vs. noise, disease vs.
non-disease, chest pain with myocardial i n f a r c t i o n vs. chest pain
without i n f a r c t i o n , blood i n stools due to malignancy vs. blood i n
stools from other conditions). This i s accuracy or correctness
r e l a t i v e to truth, as best as we can determine that truth. We can
express accuracy as c l i n i c a l s e n s i t i v i t y and s p e c i f i c i t y .
E f f i c a c y , on the other hand, i s a measure of the actual p r a c t i c a l
value of the diagnostic information or c l a s s i f i c a t i o n - how much


b e n e f i t t h e t e s t p r o v i d e s r e l a t i v e t o i t s r i s k s and c o s t s .
E v a l u a t i n g o r o p t i m i z i n g e f f i c a c y i n v o l v e s d e c i s i o n t h e o r y and
c o n s i d e r a t i o n of the complexities of c l i n i c a l u t i l i t y , r a t h e r than
j u s t accuracy.
T h i s i s p a r t o f a symposium on d e t e c t i o n l i m i t s . In this
paper I w i l l c o n s i d e r l i m i t s i n terras o f c l i n i c a l d e t e c t i o n r a t h e r
t h a n a n a l y t i c a l d e t e c t i o n . By c l i n i c a l d e t e c t i o n I mean a c c u r a c y
or the d i s c r i m i n a t i n g a b i l i t y r e f e r r e d t o i n the preceding
p a r a g r a p h . T h i s a b i l i t y o f a t e s t , e x p r e s s e d as s e n s i t i v i t y and
s p e c i f i c i t y , i s n i c e l y d e s c r i b e d and a p p r e c i a t e d u s i n g t h e
r e c e i v e r o p e r a t i n g c h a r a c t e r i s t i c (ROC) c u r v e because i t p r o v i d e s
a pure index of accuracy, of d i s c r i m i n a t i o n a b i l i t y . I t deals
w i t h s i g n a l d e t e c t i o n and t h e a b i l i t y t o d i s t i n g u i s h s i g n a l f r o m
noise. The i n d e x o f a c c u r a c y p r o v i d e d i s independent o f any
d e c i s i o n c r i t e r i o n w h i c h might be a p p l i e d o r o f any b i a s w h i c h t h e
system might have towar
d e c i s i o n aspect, which
s e p a r a t e d o u t so as n o
i n t r i n s i c a b i l i t y o f t h e t e s t t o d i s c r i m i n a t e among v a r i o u s
s t a t e s . The i n f l u e n c e o f v a r i o u s d e c i s i o n f a c t o r s ( p r e v a l e n c e ,
u t i l i t i e s ) on t h e o p e r a t i o n and u l t i m a t e e f f i c a c y o f t h e t e s t i s
a d d r e s s e d by c l i n i c a l d e c i s i o n a n a l y s i s . The f o r m a l t o o l o f
c l i n i c a l d e c i s i o n a n a l y s i s j o i n s the estimates of the
p r o b a b i l i t i e s o f t e s t outcomes ( t r u e p o s i t i v e s , f a l s e p o s i t i v e s ,
e t c . ) p r o v i d e d by ROC a n a l y s i s w i t h d e c i s i o n f a c t o r s so as t o
e s t a b l i s h t h e d e c i s i o n c r i t e r i o n f o r t e s t s and t o choose t h e s e t
and o r d e r o f d i a g n o s t i c and t h e r a p e u t i c s t e p s t o be t a k e n t o
o p t i m i z e t h e outcome i n terms o f y e a r s o f l i f e , q u a l i t y o f l i f e ,
c o s t s , resource u t i l i z a t i o n , e t c . (2-3).
The b a s i c j o b o f a c l i n i c a l l a b o r a t o r y t e s t i s t o p r o v i d e
i n f o r m a t i o n about t h e c l i n i c a l s t a t e o f p a t i e n t s f o r h e a l t h c a r e
management p u r p o s e s . The g o a l then i s t o s u b d i v i d e o r c l a s s i f y
s e e m i n g l y s i m i l a r s u b j e c t s i n t o c l i n i c a l l y r e l e v a n t management
subgroups. Suppose we a r e t a l k i n g about p e o p l e who come t o an
emergency room w i t h a c u t e c h e s t p a i n . Some w i l l t u r n o u t t o be
h a v i n g a h e a r t a t t a c k and some won't. L a b o r a t o r y t e s t s h e l p
d i v i d e o r c l a s s i f y t h o s e p a t i e n t s i n t o subgroups - t h a t i s , l a b
t e s t s h e l p t o d i s t i n g u i s h t h o s e who p r o b a b l y a r e h a v i n g a h e a r t
a t t a c k from those who a r e n ' t . The q u e s t i o n i s , what i s t h e l i m i t
of t h e a b i l i t y o f t h e t e s t t o i d e n t i f y o r d e t e c t s u b j e c t s h a v i n g a
h e a r t a t t a c k among t h o s e w i t h c h e s t p a i n ? What a r e t h e l i m i t s o f
t h e t e s t ' s powers t o d e t e c t a c c u r a t e l y t h e c l i n i c a l s t a t e o f each
i n d i v i d u a l i n t h e group? T h i s i s a s i g n a l d e t e c t i o n t h e o r y i s s u e .
Most d i a g n o s t i c t e s t s a r e i m p e r f e c t and, p a r t i c u l a r l y when we
use a b i n a r y approach - r e s u l t s a r e e i t h e r " p o s i t i v e " o r
" n e g a t i v e " - t h e r e a r e some m i s c l a s s i f i c a t i o n e r r o r s ,
i n a c c u r a c i e s . Some s u b j e c t s w i t h t h e c o n d i t i o n o f i n t e r e s t w i l l
be m i s s e d o r some w i t h o u t t h e c o n d i t i o n w i l l be m i s t a k e n l y
c o n s i d e r e d a f f e c t e d , o r b o t h w i l l happen. The a b i l i t y o f a t e s t
to properly i d e n t i f y or c l a s s i f y subjects o r conditions of
i n t e r e s t can be e x p r e s s e d as t h e s e n s i t i v i t y and s p e c i f i c i t y o f
t h e t e s t . F o r c l i n i c a l purposes t h e s e a r e d e f i n e d as f o l l o w s :
SENSITIVITY (TRUE POSITIVE RATE): F r a c t i o n o f a l l a f f e c t e d

8. ZWEIG Establishing Clinical Detection Limits of Laboratory Tests 151
subjects i n whom the test result i s p o s i t i v e ; "test p o s i t i v i t y i n

the presence of the disease." SPECIFICITY (TRUE NEGATIVE RATE):
Fraction of a l l unaffected subjects i n whom the test result i s
negative; "test negativity i n the absence of the condition."
These inaccuracies i n terms of s e n s i t i v i t y and s p e c i f i c i t y can be
s t a t i s t i c a l l y represented by the ROC curve.
This paper w i l l discuss basic test performance i n terms of
accuracy, but w i l l not deal with actual a p p l i c a t i o n of a test.
The l a t t e r involves choosing decision levels ( i . e . , reference
values, cut-offs, normal l i m i t s , etc.) and involves measures of
u t i l i t y which are beyond the scope of fundamental test
performance. I w i l l describe a set of p r i n c i p l e s or elements
important f o r evaluating test performance and comparing tests to
one another (4,5). I w i l l p a r t i c u l a r l y emphasize the power and
convenience of ROC curves, an extremely e f f e c t i v e t o o l for
assessing and comparin
usefulness of ROC curve
members of various biomedica d i s c i p l i n e years,
t o o l has received l i t t l e attention from the c l i n i c a l laboratory
community.
SiRnal/Noise Discrimination: H i s t o r i c a l Perspectives
The ROC curve apparently had i t s origins i n e l e c t r o n i c signal

detection theory. Much of t h i s arose i n the 1940's and 1950's
from analysis of radar systems. During WWII, radar operators
watched screens f o r b l i p s which might indicate enemy a i r c r a f t f o r
the purpose of deciding when to mobilize f i g h t e r squadrons to
intercept. The problem was to d i s t i n g u i s h between signals from
h o s t i l e planes and noise from clouds, flocks of b i r d s , etc.
They realized that i n interpreting the radar signals they saw
there was always a trade-off between s e n s i t i v i t y and s p e c i f i c i t y -
as the s e n s i t i v i t y increased so did the rate of f a l s e p o s i t i v e s .
That i s , i f they lowered the threshold f o r which b l i p s they
interpreted as s i g n i f y i n g enemy planes, they f a l s e l y i d e n t i f i e d
clouds and migrating b i r d s , etc., as planes more often.
S p e c i f i c i t y declined and they scrambled interceptor squadrons
unnecessarily. On the other hand, r a i s i n g the threshold f o r
c a l l i n g a b l i p " p o s i t i v e " (enemy bombers) meant not responding to
the a r r i v a l of enemy a i r c r a f t in some instances (false
negatives). They were experiencing the trade-off between
s e n s i t i v i t y and s p e c i f i c i t y inherent i n t e s t systems.
Figure 1 shows hypothetical signals and noise i n the form of
peaks. Imagine t h i s i s radar information and the r e a l planes give
peaks I, I I , and I I I . I f interceptor planes are sent up when the
signal exceeds c r i t e r i o n C, then two r e a l signals, I and I I , w i l l
be missed. However, i f c r i t e r i o n A i s used so as to catch a l l
three r e a l signals of enemy a i r c r a f t , a number of noise a r t i f a c t s
w i l l be incorrectly c l a s s i f i e d as p o s i t i v e s (false p o s i t i v e s ) .

Signal/Noise Discrimination i n the C l i n i c a l Laboratory
Figure 2 i l l u s t r a t e s t h i s i n the form of serum myoglobin

concentrations obtained 5 hours after the onset of chest pain from
patients admitted to a coronary care u n i t with the suspicion of
myocardial i n f a r c t i o n . This test has been proposed by some as a
marker f o r heart attacks. Some of these patients turned out to
have a heart attack ( s o l i d bars) and some didn't (hatched bars).
Because of the overlap between "signals" and "noise," any decision
c r i t e r i o n we choose w i l l r e s u l t i n some m i s c l a s s i f i c a t i o n s . We
could choose any of various decision l e v e l s , each giving a
d i f f e r e n t s e n s i t i v i t y / s p e c i f i c i t y combination - a l l of the
possible combinations comprising the trade-offs available with
t h i s test. This spectrum of trade-offs constitutes the detection
l i m i t of t h i s t e s t and i s represented by the ROC curve.
We have defined th
as s e n s i t i v i t y and th
as s p e c i f i c i t y and can expres percentage
decimal f r a c t i o n s . A perfect test would exhibit both a
s e n s i t i v i t y and s p e c i f i c i t y of 100% or 1.0. Tests are rarely
perfect. I t would be rather unusual f o r a test to exhibit a
s e n s i t i v i t y and a s p e c i f i c i t y of 100% at the same time. Often we
hear or read that a p a r t i c u l a r test has a p a r t i c u l a r s e n s i t i v i t y
or s p e c i f i c i t y . In r e a l i t y , as noted with radar and serum
f
myoglobin, there i s n t j u s t one s e n s i t i v i t y or s p e c i f i c i t y f o r a
t e s t , but rather a continuum of s e n s i t i v i t i e s and s p e c i f i c i t i e s .
By varying the decision l e v e l (or "decision point,"
"upper limit-of-normal," "cutoff value," "reference value," e t c . ) ,
T
any s e n s i t i v i t y from 0 to 100% can be obtained. Each of these

s e n s i t i v i t i e s w i l l have a corresponding s p e c i f i c i t y . Sensitivity
and s p e c i f i c i t y occur, then, i n p a i r s . The test's accuracy i s
r e f l e c t e d i n the pairs that can occur; not a l l pairs are possible
f o r a p a r t i c u l a r t e s t . A given test w i l l have one set of
s e n s i t i v i t y - s p e c i f i c i t y pairs i n one c l i n i c a l s i t u a t i o n , but may
have a d i f f e r e n t set of pairs when applied to another c l i n i c a l
s i t u a t i o n where the group tested i s d i f f e r e n t .
The spectrum of p a i r s exhibited by a t e s t i n a given c l i n i c a l
setting characterizes or describes the accuracy of the test.
Often test users i m p l i c i t l y assume one s e n s i t i v i t y - s p e c i f i c i t y
p a i r characterizes a t e s t because they accept a conventional,
often a r b i t r a r i l y chosen, upper-limit-of-normal as the single
correct decision l e v e l f o r that test for a l l circumstances. They
accept the corresponding s e n s i t i v i t y - s p e c i f i c i t y p a i r as the
correct one f o r the t e s t . This, however, i s actually only one of
multiple possible operating points for the t e s t . When the concept
of varying the decision l e v e l (operating point) to generate a
spectrum of s e n s i t i v i t y - s p e c i f i c i t y pairs i s understood, then the
issue becomes: How good are the pairs? Also, which p a i r ( s ) works
the best for the circumstances i n which the test i s to be used?

F i g u r e 1: Diagra
a rada
r e p r e s e n t s i g n a l s from a i r c r a f t , w h i l e a l l
o t h e r peaks r e p r e s e n t n o i s e . L i n e s A, B, C,
and D r e p r e s e n t i n c r e a s i n g d e c i s i o n l e v e l
t h r e s h o l d s , which r e s u l t s i n s u c c e s s i v e l y
lower t r u e - and f a l s e - p o s i t i v e r a t e s .
500
CD
:200
<
F 100

LU
50

20
3

> 10

D
ce
LU
CO
INDIVIDUAL PATIENTS
F i g u r e 2: Serum m y o g l o b i n c o n c e n t r a t i o n s f o r 54
p a t i e n t s w i t h chest pain admitted to a
c o r o n a r y c a r e u n i t . M y o g l o b i n was measured
5 h o u r s a f t e r the o n s e t of p a i n . Solid
bars: acute myocardial i n f a r c t .
C r o s s h a t c h e d b a r s : no i n f a r c t .

ROC Curves: Derivation
To answer these questions, we f i r s t need a way to represent and

deal with a l l these d i f f e r e n t possible operating points and t h e i r
resultant performance c h a r a c t e r i s t i c s ( s e n s i t i v i t y / s p e c i f i c t y
p a i r s ) . The ROC curve graphically displays the entire spectrum of
a given test's performance for a p a r t i c u l a r sample group of
affected and unaffected subjects. Figure 3 contains a
hypothetical frequency d i s t r i b u t i o n histogram at the top and the
and the corresponding ROC curve below. The ROC curve plots the
true p o s i t i v e (TP) rate or percentage as a function of the f a l s e
p o s i t i v e (FP) rate or percentage as the decision l e v e l i s varied.
The true p o s i t i v e rate i s the same as s e n s i t i v i t y and i s equal to
the number of affected individuals with a " p o s i t i v e " r e s u l t
divided by the t o t a l number of affected i n d i v i d u a l s . The true
p o s i t i v e rate i s also equal t 1 - f a l s negativ (FN) rate Th
f a l s e p o s i t i v e rate i s
nevertheless have a " p o s i t i v e
related to s p e c i f i c i t y , or the a b i l i t y of the test to c o r r e c t l y
i d e n t i f y unaffected individuals ( s p e c i f i c i t y = true negative (TN)
rate = number of unaffected individuals with "negative" r e s u l t s /
t o t a l number of unaffected individuals = 1 - f a l s e p o s i t i v e r a t e ) .
Both the TP and FP rates depend on the decision l e v e l chosen.
Both rates also depend on the c l i n i c a l s e t t i n g , as r e f l e c t e d by
the study population chosen. The FP rate i s influenced by the
type of nondiseased subjects included i n the study group. I f , for
example, the nondiseased subjects are a l l healthy blood donors who
are free of any signs or symptoms of disease, the test may appear
to have a much lower rate than i f the nondiseased subjects are
persons who c l i n i c a l l y resemble those who a c t u a l l y have the
disease. Like the FP rate, the TP rate also depends on the study
group. A test used to detect cancer may have a higher TP rate
when applied to patients who have active or advanced disease than
when applied to patients having stable or limited disease. This
dependence of TP and FP rates on the study population i s the
reason why an ROC curve must be generated f o r each c l i n i c a l
situation.
Each point on the ROC curve represents a p a i r of true and
f a l s e p o s i t i v e rates corresponding to some decision l e v e l . In
Figure 3 , the l e f t hand curve of the frequency histogram (top)
represents results from unaffected individuals and the right hand
curve i s derived from affected individuals. The ROC curve i s
derived from the data i n the frequency histogram, so the f i r s t
step i s to obtain the t e s t results from both the affected group
and the unaffected group. True p o s i t i v e rates are calculated
using the results from the affected i n d i v i d u a l s , while f a l s e p o s i -
t i v e rates are generated from the unaffected individuals' data.
The ROC curve i s constructed by varying the decision l e v e l from
the highest test r e s u l t down to zero, r e s u l t i n g i n true and f a l s e
p o s i t i v e rates which vary continuously. The decision l e v e l at
point a i n Figure 3 i s higher than any observed results (see top),
so at that decision l e v e l none of the r e s u l t s are " p o s i t i v e " and
both true and f a l s e p o s i t i v e rates are zero (see bottom). As the

unaffected
False Positive Rate (%)
Figure 3: Top: Hypothetical frequency d i s t r i b u t i o n

curve. Bottom: Receiver operating
c h a r a c t e r i s t i c (ROC) curve corresponding to
data i n top panel, generated by varying the
decision l e v e l and then p l o t t i n g the
r e s u l t i n g pairs of true and f a l s e p o s i t i v e
rates. Arrows at a to e mark points
corresponding to decision l e v e l s i n top
panel. The curve from c to d d e s c r i b e s t h e
test's performance i n the c r u c i a l overlap
region.

decision l e v e l i s lowered from a to b, some of the affected

individuals have p o s i t i v e results but none of the unaffected
individuals do, so the true p o s i t i v e rate r i s e s while the f a l s e
p o s i t i v e rate remains zero. Point c shows the highest true
p o s i t i v e rate achievable (with t h i s data) with the f a l s e p o s i t i v e
rate s t i l l at zero. This i s the edge of the overlap region (c to
d). At c the ROC curve leaves the Y axis because i f the decision
l e v e l i s lowered any further, some unaffected individuals have
f a l s e l y p o s i t i v e r e s u l t s . At decision l e v e l d, a l l affected
individuals have p o s i t i v e test r e s u l t s , so the true p o s i t i v e rate
reaches 100%, at the expense of some percentage of f a l s e
p o s i t i v e s . This i s the other edge of the c r u c i a l overlap region.
The portion of the curve from c to d (where i t has l e f t the Y
axis but not yet intercepted the true p o s i t i v e = 100% horizontal
l i n e ) describes the overlap region. From decision l e v e l d to e,
f a l s e p o s i t i v e rates increas
unaffected individuals
ROC Curves: Interpretation
The complete ROC curve summarizes the c l i n i c a l accuracy of the

test by displaying the paired true and f a l s e p o s i t i v e rates for
a l l possible decision l e v e l s . Good c l i n i c a l performance of a test
i s characterized by a high true p o s i t i v e rate and a low f a l s e
p o s i t i v e rate. Accordingly, as test performance improves, the ROC
curve w i l l move upward (toward higher true p o s i t i v e rates) and to
the l e f t (toward lower f a l s e p o s i t i v e r a t e s ) . A perfect test
would achieve a 100% true p o s i t i v e rate with no f a l s e p o s i t i v e s .
Thus, i t s ROC curve would r i s e v e r t i c a l l y to the (0,100) point i n
the upper l e f t comer and then move h o r i z o n t a l l y to the right
along the horizontal l i n e representing true p o s i t i v e rate = 100%
to the (100,100) point i n the upper right corner. Conversely, for
a c l i n i c a l l y useless t e s t , which gives s i m i l a r results for
subjects with and without the condition, the true and f a l s e
p o s i t i v e rates would be i d e n t i c a l for any given decision l e v e l .
Therefore, the ROC curve would be a diagonal between the lower
l e f t and upper r i g h t corners, representing the l i n e where the true
p o s i t i v e rate always equals the f a l s e p o s i t i v e rate.
Because the curve i s usually above the diagonal, i t starts out
at the lower l e f t with the TP rate ( s e n s i t i v i t y ) increasing faster
than the f a l s e p o s i t i v e rate. At some point the slope begins to
f a l l and the f a l s e p o s i t i v e rate starts increasing faster than the
true p o s i t i v e rate - i n other words, gains i n s e n s i t i v i t y come at
the cost of increasingly larger costs in terms of n o n s p e c i f i c i t y .
This imposes a p r a c t i c a l l i m i t on the usable s e n s i t i v i t y of the
test - where that l i m i t i s depends on the r e l a t i v e u t i l i t y or
benefits and the costs of true and f a l s e r e s u l t s and gets us
beyond detection and into decision issues.
The ROC curve can also be constructed as a p l o t of true
p o s i t i v e rate ( s e n s i t i v i t y ) versus true negative rate
( s p e c i f i c i t y ) instead of versus f a l s e p o s i t i v e rate

( 1 - s p e c i f i c i t y ) . T h i s produces a m i r r o r image of the c u r v e shown

i n F i g u r e 3, f l i p p i n g t h e c u r v e t o the r i g h t s i d e w i t h the p e r f e c t
p o i n t b e i n g the upper r i g h t hand c o r n e r i n s t e a d of the upper l e f t
hand c o r n e r .
The ROC c u r v e , t h e n , p r o v i d e s a comprehensive p i c t u r e of t h e
t e s t ' s accuracy at a l l p o s s i b l e operating p o i n t s ( d e c i s i o n
l e v e l s ) . I t does t h i s w i t h o u t the need t o choose a d e c i s i o n l e v e l
o r e s t a b l i s h a normal range i n advance.
Comparing T e s t s
B e s i d e s b e i n g v a l u a b l e i n e v a l u a t i n g a s i n g l e t e s t by
d e m o n s t r a t i n g the complete spectrum of i t s i n t r i n s i c performance,
t h e ROC c u r v e i s e x t r e m e l y u s e f u l i n comparing t e s t s t o one
a n o t h e r . Even i f we a r e e v a l u a t i n g o n l y a s i n g l e new t e s t ,
comparisons t o e x i s t i n
process. ROC c u r v e s p r o v i d
d e m o n s t r a t i n g the r e l a t i v y multipl , comparing
them a t e v e r y TP r a t e by p l o t t i n g the ROC c u r v e s f o r a l l the t e s t s
on t h e same graph. I f the ROC c u r v e f o r one t e s t i s u n i f o r m l y
above and t o the l e f t o f the ROC c u r v e f o r a second t e s t , t h e
f i r s t t e s t w i l l have a lower FP r a t e t h a n t h e second t e s t has f o r
any g i v e n TP r a t e .
The ROC c u r v e s of F i g u r e 4 i l l u s t r a t e t h e a m b i g u i t y i n v o l v e d
i n comparing t e s t s a t j u s t one d e c i s i o n l e v e l o r o p e r a t i n g p o i n t .
C o n s i d e r the case i n w h i c h t e s t A has a TP r a t e of 98% and a FP
r a t e of 30%, w h i l e t e s t has a TP r a t e of 70% and a FP r a t e of
2%. I f the c l i n i c a l performance of the two t e s t s were e q u i v a l e n t ,
t h e y would s h a r e a s i n g l e ROC c u r v e . This s i t u a t i o n i s
i l l u s t r a t e d i n F i g u r e 4, l e f t . T e s t c o u l d have a c h i e v e d t h e
same TP and FP r a t e s as t e s t A i f a d i f f e r e n t d e c i s i o n l e v e l had
been used. I n f a c t e i t h e r t e s t c o u l d have a c h i e v e d any of t h e
p a i r s of TP and FP r a t e s on the common ROC c u r v e s i m p l y by
c h a n g i n g the d e c i s i o n l e v e l . Thus, the two t e s t s may i n f a c t
s h a r e a s i n g l e ROC c u r v e but i n i t i a l l y appear t o p e r f o r m
d i f f e r e n t l y because t h e two d e c i s i o n l e v e l s used p l a c e the t e s t s
a t d i f f e r e n c e p o i n t s on the c u r v e , i . e . , t h e o p e r a t i n g c o n d i t i o n s
were not comparable. On t h e o t h e r hand, t h e two t e s t s may
a c t u a l l y perform very d i f f e r e n t l y , with t e s t c l e a r l y s u p e r i o r ,
as i l l u s t r a t e d i n F i g u r e 4, c e n t e r . R e g a r d l e s s of the d e c i s i o n
l e v e l chosen f o r t e s t A, i t can not a c h i e v e a TP r a t e of 70% w i t h
a FP r a t e of o n l y 2%, as d i d t e s t B. I n f a c t , when t e s t A*s TP
r a t e i s 70%, i t s FP r a t e i s 10%. S i m i l a r l y , t h e t r u e - and f a l s e
p o s i t i v e r a t e s g i v e n o r i g i n a l l y would be e q u a l l y c o n s i s t e n t w i t h
t h e s i t u a t i o n shown i n F i g u r e 4, r i g h t , where t e s t A i s c l e a r l y
s u p e r i o r . These examples i l l u s t r a t e how t h e use of ROC c u r v e s
a v o i d s the a m b i g u i t y w h i c h may o c c u r when t e s t s a r e compared u s i n g
o n l y one d e c i s i o n l e v e l f o r each.

en
oo
100

~~ to

- 30
20 M
H
h- M
10
10 20 30 40 SO 60 70 tO 90 10 20 30 40 SO 60 70 tO 90 100 10 20 30 40 SO 0 70 10 *0

FALSE POSITIVE RATE () FALSE POSITIVE RATE IX) FALSE POSITIVE RATE (X)
*
F i g u r e 4. H y p o t h e t i c a l r e c e i v e r o p e r a t i n g c h a r a c t e r i s t i c (ROC) c u r v e s showing

t h r e e p o s s i b l e r e l a t i o n s between t e s t s A and B. I n each c a s e , t e s t A e x h i b i t s >
1
a t r u e p o s i t i v e r a t e o f 98% and a f a l s e p o s i t i v e r a t e o f 30%, w h i l e t e s t r
e x h i b i t s a t r u e p o s i t i v e r a t e o f 70% and a f a l s e p o s i t i v e r a t e o f 2%. L e f t
p a n e l : Both t e s t s have i d e n t i c a l ROC c u r v e s , and t h u s , e q u i v a l e n t d i a g n o s t i c
>
accuracy. Middle panel: Test has a b e t t e r ROC c u r v e . Right panel: Test 1
r

A has a b e t t e r c u r v e .
ROC Curves: Application
Figures 5 and 6 are examples of real ROC curves and i l l u s t r a t e how

the ROC curve can represent individual t e s t accuracy as well as
compare the accuracy of multiple tests to one another. Four
analytes were measured. Creatine kinase (CK) i s a serum enzyme,
found primarily i n heart and other muscles, which has been used
f o r some years as an early marker for necrosis. Peak serum
concentrations usually occur within the f i r s t 12-24 hours a f t e r
the onset of i n f a r c t i o n . CK-MB i s an isoenzyme of CK which i s
more s p e c i f i c f o r heart muscle than i s t o t a l CK and thus has
become popular i n the l a s t 10 years. CK-BB, another isoenzyme of
CK found i n the heart, has also been examined as a possible marker
f o r myocardial i n f a r c t i o n . Myoglobin, a heme containing protein
found i n muscle, i s released into the serum with muscle injury.
Serum concentrations of myoglobi t r i s e a r l i e tha C
i n patients with myocardia
a f t e r the onset of ches
these four markers of myocardial injury i n patients suspected of
having a heart attack sampled 8 hours a f t e r the onset of chest
pain. Myoglobin occupies the left-most p o s i t i o n of the tests, and
achieves the best r a t i o of true positives to f a l s e positives, with
good absolute s e n s i t i v i t y (high true p o s i t i v e rate) and
s p e c i f i c i t y (low f a l s e p o s i t i v e rate) simultaneously. From the
ROC curve, one can make two judgements. F i r s t , myoglobin achieves
the best accuracy of the four tests. Second, myoglobin probably
has potential as a early marker of myocardial i n f a r c t i o n because
i t ' s ROC curve l i e s quite close to the i d e a l location, the upper
l e f t hand corner. This indicates that i t can achieve high true
p o s i t i v e and low f a l s e p o s i t i v e rates at the same time. How best
to use this test c l i n i c a l l y and which decision l e v e l ( i . e . , where
on the ROC curve to operate) to select requires c l i n i c a l decision
analysis with consideration of the costs of f a l s e results, the
alternative tests or procedures available, the costs of the
alternatives, and the u t i l i t i e s of the various possible outcomes
(2,3). The ROC curve displays the spectrum of
s e n s i t i v i t y / s p e c i f i c i t y pairs achievable; these pairs are the raw
data needed to make the selection of decision l e v e l .
In Figure 6, the patients are sampled at 18 hours after the
onset of chest pain. Myoglobin's accuracy has decreased while
that of the three other tests has markedly increased to a
close-to-perfect l e v e l . This r e f l e c t s the fact that the serum
concentration of myoglobin i s not increased as much or as often at
18 hours compared to 8 hours after the onset of pain. Therefore,
i t i s not as good at discriminating between patients having and
not having an i n f a r c t i o n . CK and i t s isoenzymes, on the other
hand, are near peak concentrations i n those patients with i n f a r c t s
and lower i n those without i n f a r c t s , and thus are very accurate i n
discriminating. In t h i s study, the "true" diagnosis or gold
standard was established by review of electrocardiographic data,
c l i n i c a l course, and serum lactate dehydrogenase isoenzymes, as

10 20 30 40 50 70 *0 90 100
F A L S E P O S I T I V E RfiTE (%)
F i g u r e 5: ROC c u r v e s o f 4 serum t e s t s 8 hours a f t e r

the onset of chest p a i n i n p a t i e n t s
suspected of having a myocardial i n f a r c t i o n .
CK = c r e a t i n e k i n a s e ; CK-BB = " b r a i n "
isoenzyme o f CK; CK-MB = " m y o c a r d i a l "
isoenzyme o f CK.

100 .
90
so +
Lu 70
H
<E
d 60
Lu
> 50
40 -

UJ 30
MYOGLOBIN
(
20 - CK-BB
CK-nB
10 TOTAL CK
0 +- 4- +- -f- -4-
10 20 30 40 50 60 70 SO 90 100
FALSE POSITIVE RATE (%)
F i g u r e 6: ROC c u r v e s o f 4 serum t e s t s 18 hours a f t e r

the o n s e t o f c h e s t p a i n i n p a t i e n t s
suspected of having a myocardial
i n f a r c t i o n . A b b r e v i a t i o n s a r e same as f o r
F i g u r e 5.

well as scintigraphic findings where available. To avoid

introduction of bias, the c l a s s i f i c a t i o n of patients was made
without consideration of the results of any of the four tests
being evaluated.
A given study provides an estimate of the ROC curve for that
t e s t and patient population. The confidence l i m i t s around the ROC
curve can be calculated (8,9). Furthermore, the area under the
ROC curve can be calculated f o r each test so as to derive a
quantitative index of the test's individual accuracy and i t s
r e l a t i o n to the other tests being evaluated (8,9).
ROC curves can also be used to examine the impact of
a n a l y t i c a l improvements on c l i n i c a l accuracy. Figure 7 shows
d i s t r i b u t i o n s of test results and the corresponding ROC curves,
based on simulated data. For both the affected and the unaffected
patients, the b i o l o g i c a l v a r i a b i l i t y of the marker being measured
has a standard deviatio
have a mean test r e s u l
a mean test result of 12 a n a l y t i c a imprecisio
SD of 4 units, there i s considerable overlap i n test results
between the affected and unaffected patients. The corresponding
ROC curve shows the poor c l i n i c a l performance of the test.
If an improved a n a l y t i c a l system reduces the imprecision of
the measurement from 4 to 2 units, the overlap i n test results i s
considerably reduced. The dramatic s h i f t of the ROC curve upward
and to the l e f t r e f l e c t s the improved c l i n i c a l performance of the
test.
If the a n a l y t i c a l imprecision i f again halved, reducing i t s
standard deviation from 2 to 1, another s i g n i f i c a n t improvement i n
c l i n i c a l performance occurs. In this example, i n which the
b i o l o g i c a l overlap between the two groups of patients was small,
the precision of the a n a l y t i c a l system became the p r i n c i p a l factor
i n determining the c l i n i c a l performance of the test; substantial
improvements i n c l i n i c a l accuracy occurred as the a n a l y t i c a l
p r e c i s i o n improved.
In contrast, Figure 8 shows the s i t u a t i o n i n which the
b i o l o g i c a l overlap i s greater. In t h i s example, the b i o l o g i c a l
v a r i a t i o n i n each group has an SD of 4 u n i t s , r e s u l t i n g i n
considerable i n t r i n s i c overlap i n the test results of the two
groups. The figure shows t h i s extensive overlap and the poor ROC
curve for an a n a l y t i c a l SD of 4. Decreasing the a n a l y t i c a l
imprecision (from 4 to 2 to 1) provides only a minor improvement
i n c l i n i c a l accuracy. Thus, when the b i o l o g i c a l overlap of the
two groups i s large, even severalfold improvements i n a n a l y t i c a l
p r e c i s i o n may have l i t t l e e f f e c t on the c l i n i c a l accuracy of the
t e s t , as r e f l e c t e d i n the ROC curve.
P r i n c i p l e s of Test Evaluation
Once we have the basic performance data describing detection and

the l i m i t s of detection as represented by the ROC curve, then we
can go on to decision analysis. This involves structuring the
c l i n i c a l problem i n the form of a decision tree, estimating
u t i l i t i e s and costs of various outcomes, choosing decision levels

ANALYTICAL SD=V ANALYTICAL SD=2
BIOLOGICAL SD-1 BIOLOGICAL SD-I
s
Lu
Lu
Id
>
(L
' \/ \
8 12 16 20 2t 28 32
i l \
TEST RESULT TEST RESULT

F i g u r e 7. Frequency d i s t r i b u t i o n and c o r r e s p o n d i n g ROC c u r v e s f o r a t e s t
h a v i n g a b i o l o g i c a l SD o f 1 and a n a l y t i c a l SD o f 1, 2, o r 4. The two
peaks i n t h e f r e q u e n c y d i s t r i b u t i o n s r e p r e s e n t t e s t r e s u l t s from d i s e a s e d

and n o n d i s e a s e d p o p u l a t i o n s . C o n t i n u e d on n e x t page.
se
>
>
w
>

q
M

r

ANALYTICAL SD=V ANALYTICAL SD=2
BIOLOGICAL BIOLOGICAL SD-V
TEST RESULT TEST RESULT

F i g u r e 8. Frequency d i s t r i b u t i o n and c o r r e s p o n d i n g ROC c u r v e s f o r a t e s t
h a v i n g a b i o l o g i c a l SD o f 4. Other f e a t u r e s same as F i g u r e 7. Continued
on next page.

1
H
M
H
M

for using the t e s t . However, obtaining v a l i d data f o r the ROC

curve i n the f i r s t place requires attention to several
common-sense p r i n c i p l e s which are suprisingly often overlooked.
Table I has a l i s t or recipe for designing a good study to
evaluate the detection power of a test. The ideal study i s
prospective and i s usually harder, longer and more expensive than
the type of evaluation commonly done, but an "inexpensive'*
c l i n i c a l evaluation may prove more costly i n the long run i f i t s
erroneous conclusions lead to improper test u t i l i z a t i o n or
improper patient management.
Table I. P r i n c i p l e s of a Good Evaluation of a Laboratory Test

1. DEFINE CLINICAL QUESTION TEST WILL BE USED FOR
2. SELECT APPROPRIATE SUBJECTS TO STUDY
3. CLASSIFY
4. PERFORM AL
5. EVALUATE
The f i r s t and most important element on this l i s t i s defining

s p e c i f i c a l l y and c a r e f u l l y the c l i n i c a l question or problem at
which the test i s to be directed. I t ' s not enough to say "Let's
look at this test for p r o s t a t i c cancer or coronary artery disease
and see how well i t does." We need to define precisely what
question of relevance to patient management i s being addressed and
how that test w i l l be used i n practice. Do we want to screen
large numbers of people for cancer or use the test to e s t a b l i s h
the stage of cancer once we know i t ' s there, or do we want to
predict response to a p a r t i c u l a r therapy, or assess response to a
p a r t i c u l a r therapy? I t may provide a l l these functions but with
varying effectiveness and requiring d i f f e r i n g decision levels.
Each of these roles must be evaluated separately because the
populations are d i f f e r e n t , conditions are d i f f e r e n t , goals are
d i f f e r e n t , and ROC curves may be d i f f e r e n t .
If you think about these issues, c a r e f u l l y and s p e c i f i c a l l y
defining what you are trying to establish, the rest starts f a l l i n g
into place.
The second element i s selecting appropriate subjects. Once
you have defined the question, you've pointed the way toward the
proper subjects. I f you want to use a tumor marker to i d e n t i f y
colon cancer among middle aged people with bowel obstruction,
occult blood loss, or unexplained anemia, then you need to look at
the test performance i n that group of subjects. Healthy young
people aren't relevant and neither i s a reference range based on
them. There's no point i n doing conventional normal ranges i f
healthy young volunteers aren't the ones f o r whom the test i s
intended.
Number three concerns establishing the true diagnosis: Once
you've got a group of people with bowel signs or symptoms
suggestive that cancer i s possible, then you must separate them
into 2 groups, those who r e a l l y do have carcinoma of the colon and
those who don't. This provides a gold standard f o r calculation of

TP r a t e s , FP r a t e s , e t c . T h i s d i a g n o s i s needs t o be a c c u r a t e as
w e l l as independent of a l l t e s t s b e i n g e v a l u a t e d . To t h e e x t e n t
t h a t e i t h e r a c c u r a c y o r independence i s l a c k i n g , t h e r e s u l t s o f
t h e e v a l u a t i o n w i l l be b i a s e d and m i s l e a d i n g . C o n s i d e r the
h y p o t h e t i c a l s i t u a t i o n i n F i g u r e 9. The c l i n i c a l q u e s t i o n i s ,
"Has t h i s p a t i e n t p r e s e n t i n g a t the emergency room w i t h an a c u t e
p s y c h i a t r i c d i s o r d e r used m a r i j u a n a r e c e n t l y ? " The r o u t i n e t e s t
i s s e n s i t i v e enough t o d e t e c t o n l y 70% of t h e r e c e n t drug u s e r s ;
30% of t h e m a r i j u a n a u s e r s have f a l s e l y n e g a t i v e r e s u l t s . The
r o u t i n e t e s t a l s o s u f f e r s from v a r i o u s i n t e r f e r e n c e s , l e a d i n g t o
f a l s e p o s i t i v e r e s u l t s i n 30% of n o n - u s e r s . T e s t I r e p r e s e n t s a
new t e s t which i s b e i n g e v a l u a t e d . I n a c t u a l i t y i t m a n i f e s t s
e x c e l l e n t s e n s i t i v i t y and s p e c i f i c i t y , g i v i n g p o s i t i v e r e s u l t s i n
a l l r e c e n t m a r i j u a n a u s e r s and n e g a t i v e r e s u l t s i n a l l n o n - u s e r s .
I f , however, i n s t e a d of i n d e p e n d e n t l y and a c c u r a t e l y d e t e r m i n i n g
t h e drug-use s t a t u s of each p a t i e n t t h e p a t i e n t s a r e s i m p l y
c l a s s e d as u s e r s o r non-user
r e s u l t s , T e s t I w i l l appea
o f t h e p a t i e n t s . I n t h i s case a p e r f e c t t e s t appears t o p e r f o r m
p o o r l y s i m p l y because t h e c l i n i c a l q u e s t i o n was n o t answered
a c c u r a t e l y f o r each p a t i e n t ; i . e . , the " g o l d s t a n d a r d " used f o r
comparison was inadequate.
The o p p o s i t e b i a s can a l s o r e s u l t from use of inadequate g o l d
standards. T e s t I I i n F i g u r e 9 performs even more p o o r l y t h a n t h e
r o u t i n e t e s t , y i e l d i n g f a l s e n e g a t i v e r e s u l t s i n 40% of the
m a r i j u a n a u s e r s and f a l s e p o s i t i v e r e s u l t s i n 40% of the
n o n - u s e r s . I f , however, t h e r o u t i n e t e s t ' s r e s u l t s a r e a c c e p t e d
as c o r r e c t and T e s t I I i s judged on t h i s b a s i s , T e s t I I w i l l
appear t o m i s c l a s s i f y o n l y 10% of the p a t i e n t s and w i l l have a
b e t t e r apparent performance than T e s t I !
T h i s can o c c u r i n s e v e r a l ways i n c l i n i c a l p r a c t i c e . In
e v a l u a t i n g a t e s t f o r a c u t e m y o c a r d i a l i n f a r c t i o n , i f the p a t i e n t s
a r e c l a s s i f i e d on t h e b a s i s of EKG d a t a a l o n e o r even a
c o m b i n a t i o n of h i s t o r y , EKG f i n d i n g s and some c a r d i a c enzyme
r e s u l t s (a " r o u t i n e workup"), t h e d i a g n o s i s may s t i l l be
i n a c c u r a t e and, t h u s , d i s t o r t the apparent performance of t h e new
t e s t . I n t h e case of a cancer tumor maker, i f the g o l d s t a n d a r d
( d i a g n o s i s o r s t a g i n g , e t c . ) i s based upon c l i n i c a l f i n d i n g s
r a t h e r than s u r g i c a l and/or t i s s u e d a t a , t h e n t h e g o l d s t a n d a r d
may be i n a c c u r a t e and b i a s t h e apparent v a l u e of t h e marker. I f
an a m n i o t i c f l u i d marker f o r f e t a l l u n g m a t u r i t y i s compared t o an
e x i s t i n g i m p e r f e c t marker, then even i f t h e new marker i s p e r f e c t ,
i t w i l l appear i m p e r f e c t . The g o l d s t a n d a r d a g a i n s t which t h e new
marker s h o u l d be compared i s the a c t u a l p r e s e n c e o r absence of
r e s p i r a t o r y d i s t r e s s syndrome i n those newborns d e l i v e r e d w i t h i n a
s h o r t time of measurement of the marker.
Because t h e v a l i d i t y of a c l i n i c a l e v a l u a t i o n ' s c o n c l u s i o n s i s
c r i t i c a l l y dependent on t h e a c c u r a t e d e t e r m i n a t i o n of t h e answer
t o t h e c l i n i c a l q u e s t i o n f o r each s u b j e c t , r o u t i n e c l i n i c a l
diagnoses are l i k e l y t o be inadequate f o r t e s t e v a l u a t i o n
s t u d i e s . D e f i n i t i v e d e t e r m i n a t i o n of a p a t i e n t ' s t r u e c l i n i c a l
subgroup may r e q u i r e such p r o c e d u r e s as b i o p s y , s u r g i c a l
e x p l o r a t i o n , a u t o p s y e x a m i n a t i o n , a n g i o g r a p h y , o r l o n g term
f o l l o w - u p of response t o t h e r a p y and c l i n i c a l outcome.

ZWEIG Establishing Clinical Detection Limits of Laboratory Tests
Patients Who Have Patients Who Have Not

Used Marijuana
Routine Test Test Routine Test Test

Test I II Test 1 II
100 100
- FN FN FP FP
= TP: 80 II0H11
60 60
I 40 ~ TP 40 " TN TN
TP
20 20
0
FN = False Negative Results FP = False Positive Results

TP = True Positive Results TN = True Negative Results
F i g u r e 9: H y p o t h e t i c a l performances o f t h r e e t e s t s f o r
m a r i j u a n a u s e i n two subgroups o f p a t i e n t s ,
one w h i c h has used m a r i j u a n a r e c e n t l y and
one w h i c h has n o t . Assumes t h a t t h e r o u t i n e
t e s t g i v e s c o r r e c t r e s u l t s i n 70% o f
subjects.

The next item i s performing a l l tests being evaluated on a l l

the subjects being used. This may sound reasonable but i s very
often overlooked. I f the specimens or subjects aren't i d e n t i c a l
for a l l tests being examined, observed differences i n test
performance could simply be r e f l e c t i o n s of differences i n the
subjects rather than true differences i n performance.
The last element i s evaluating and comparing tests using ROC
curves, extensively discussed above. The ROC analysis i s a
powerful tool which provides a pure index of accuracy, of
discrimination c a p a b i l i t y , c l e a r l y describing the limits of
c l i n i c a l detection possible for a given test i n a given c l i n i c a l
setting. Adherence to the recipe i n Table I, including ROC
analysis, should maximize the likelihood of obtaining a v a l i d
assessment of laboratory test accuracy.
Literature Cited
1. Swets, J.., and Pickett, Diagnosti
Systems. Methods from Signal Detection Theory; Academic
Press: New York, 1982; Chapter 1.
2. McNeil, B.J.; Keeler, E.; Adelstein, S.J. N. Engl. J. Med.
1975, 293, 211-215.
3. Weinstein, C.; Feinberg, H.V. Clinical Decision Analysis;
W.B. Saunders Co.: Philadelphia, 1980.
4. Zweig, M.H.; Robertson, E.A. Clin. Chem. 1982, 28, 1272-1276.
5. Robertson, ..; Zweig, M..; Van Steirteghem, A.C. Amer. J.
Clin. Pathol. 1983, 79, 78-86.
6. Metz, C.E. Semin. Nucl. Med. 1978; 8, 283-298.
7. Turner, D.A. J. Nucl. Med. 1978, 19, 213-220.
8. Beck, J.R.; Shultz, E.K. Arch. Pathol. Lab. Med. 1986, 110,
13-20.
9. McNeil, B.J.; Hanley, J.A. Med. Dec. Making. 1984, 4,
137-150.

Chapter 9
Perspectives on Detection Limits

for Nuclear Measurements in Selected National
and International Programs
1 2
Lloyd A. Currie and Robert M. Parr
1
2
International Atomic Energy Agency, A-1400 Vienna, Austria
Issues involving the definition and practical significance

of Detection Limits are discussed in the light of US and
international programs in which the concept plays a
central role. The US program relates to the "Lower Limit
of Detection" (LLD) which forms a part of the Technical
Specifications for nuclear power reactors, as required by
the US Nuclear Regulatory Commission for measurements of
effluent and environmental radioactivity. The programs of
the International Atomic Energy Agency, which are oriented
toward coordinated research and technical cooperation,
similarly require common understanding and use of the
"Limit of Detection" (LOD) as a practical and meaningful
performance characteristic for measurements of trace
elements in bioenvironmental matrices. Efforts to meet
the needs of these two programs to formulate realistic and
practicable detection limits will be reviewed, with
special focus on problems-in-common such as the treatment
of the blank, decision criteria, algorithm and assumption
dependence, and the reporting of subliminal results.
Detection limits have p r a c t i c a l s i g n i f i c a n c e i n a number o f
important s o c i e t a l contexts, where measurement p r o c e s s e s must
p o s s e s s adequate d e t e c t i o n c a p a b i l i t y t o meet s p e c i f i c d i a g n o s t i c ,
r e g u l a t o r y , o r r e s e a r c h needs. Two such c o n t e x t s , where s p e c i f i c
r e q u e s t s were made t o f o r m u l a t e m e a n i n g f u l and r e l i a b l e approaches
to d e t e c t i o n l i m i t s , i n v o l v e d : a) t h e r e q u i r e m e n t s o f t h e U. S.
N u c l e a r R e g u l a t o r y Commission [NRC] f o r t h e d e t e c t i o n o f e f f l u e n t
and e n v i r o n m e n t a l r a d i o a c t i v i t y , and b) t h e r e q u i r e m e n t s o f t h e
I n t e r n a t i o n a l Atomic Energy Agency [IAEA] f o r t h e d e t e c t i o n o f t r a c e
elements i n a wide range o f b i o e n v i r o n m e n t a l matrices. The NRC
p r o j e c t r e l a t e d d i r e c t l y t o t h e f o r m a l r e g u l a t o r y r e q u i r e m e n t , as
set f o r t h i n the t e c h n i c a l s p e c i f i c a t i o n s f o r operating nuclear
power r e a c t o r s , t h a t t h e r a d i o a c t i v i t y measurement p r o c e s s e s have
d e t e c t i o n l i m i t s meeting s p e c i f i e d s t a n d a r d s f o r t h e purpose o f
p r o t e c t i n g t h e p u b l i c from e x c e s s i v e r e l e a s e s . T y p i c a l r e q u i r e m e n t s
0097-6156/88/0361-0171 $06.75/0

f o r t h e "Lower L i m i t o f D e t e c t i o n " [LLD] a r e g i v e n i n T a b l e I (1) .

The IAEA e f f o r t was f o c u s s e d on measurements o f e s s e n t i a l and t o x i c
t r a c e elements i n b i o e n v i r o n m e n t a l samples g l o b a l l y , as found i n
t h e i r l a b o r a t o r y i n t e r c o m p a r i s o n and c o o r d i n a t e d r e s e a r c h programs.
D e t e c t i o n c a p a b i l i t i e s i n t h i s case r e l a t e d i r e c t l y t o t h e a b i l i t y
to d e t e c t d e f i c i e n c i e s ( o f e s s e n t i a l elements) o r e x c e s s e s ( o f t o x i c
elements) i n foods o r b i o l o g i c a l samples from d i f f e r e n t g e o g r a p h i c a l
regions. F i g u r e 1, w h i c h i s drawn from t h e IAEA's c u r r e n t l i s t o f
a v a i l a b l e r e f e r e n c e m a t e r i a l s , i n d i c a t e s some t y p i c a l m a t r i c e s and
elements o f i n t e r e s t ( 2 . 3 ) . Concentrations v a r i e d widely with
element and m a t r i x ; by way o f i l l u s t r a t i o n , c e r t i f i e d v a l u e s on a
d r y w e i g h t b a s i s i n sample H-8 (horse k i d n e y ) ranged from 0.91 mg/kg
f o r Hg t o 12.6 g/kg f o r C I .
The two p r o j e c t s shared some f e a t u r e s i n common. R e l i a b l e and
comparable measures o f d e t e c t i o n c a p a b i l i t i e s were r e q u i r e d on a
c o n t i n u i n g b a s i s among many l a b o r a t o r i e s and they h a d t o be d e v e l
oped i n a manner t h a t woul
a b r o a d range o f a n a l y t e
matrices. D e t e c t i o n l i m i t s had t o be a b s o l u t e i n t h e sense t h a t
a c t u a l c o n c e n t r a t i o n o r r a d i o a c t i v i t y l e v e l s may be l i n k e d t o
s p e c i f i c n u t r i t i o n a l o r h e a l t h e f f e c t s . A t t h e o u t s e t o f each study
it was o b s e r v e d t h a t v a r y i n g d e f i n i t i o n s and e v a l u a t i o n s o f
d e t e c t i o n l i m i t s were i n use, w i t h t h e r e s u l t t h a t s t a t e d d e t e c t i o n
l i m i t s were n o t o n l y non-comparable b u t i n some c a s e s i n e r r o r by
o r d e r s o f magnitude. Such d i s c r e p a n c i e s a r e a l r e a d y most s e r i o u s
from a s c i e n t i f i c o r m e t r o l o g i c a l s t a n d p o i n t , b u t they invite
dangerous m i s i n t e r p r e t a t i o n and c o n f u s i o n when v i e w e d by t h e l a y
p u b l i c o r i n t e r p r e t e d by s p e c i a l i n t e r e s t groups.
Approach t o NRC and IAEA S t u d i e s
I n i t i a l States. An o u t l i n e o f t h e approach t a k e n f o r each study,

t o g e t h e r w i t h an i n d i c a t i o n o f t h e i n i t i a l d e t e c t i o n l i m i t d e f i n i -
t i o n s i n use by t h e two o r g a n i z a t i o n s i s g i v e n i n T a b l e I I . A p a r t
from t h e d i f f e r e n c e i n t e r m i n o l o g y - - Lower L i m i t o f D e t e c t i o n
[LLD], and L i m i t o f D e t e c t i o n [LOD] -- t h e f o r m u l a t i o n s d i f f e r i n
c o e f f i c i e n t s and i n u n d e r l y i n g p r i n c i p l e s . ( F o r t h e purposes o f
t h i s c h a p t e r we s h a l l r e p r e s e n t t h e d e t e c t i o n l i m i t by t h e s i n g l e
symbol Lp . ) The i n i t i a l NRC e x p r e s s i o n e x p l i c i t l y r e c o g n i z e s t h e
two h y p o t h e s i s t e s t i n g e r r o r s , (5%) f a l s e p o s i t i v e s and f a l s e
n e g a t i v e s , though w i t h a t r i v i a l r o u n d i n g e r r o r [4.66 r a t h e r than
4.65]. The i n i t i a l IAEA f o r m u l a t i o n , w h i c h was t a k e n from K e i t h
(4), e x p l i c i t l y t a k e s i n t o account o n l y t h e f a l s e p o s i t i v e e r r o r ,
but a t a g r e a t l y reduced ( 1 - s i d e d ) s i g n i f i c a n c e l e v e l (as l i t t l e as
0.13% i n contrast to 5%). T h i s i s a problem: ignoring false
n e g a t i v e s does n o t make them go away! I n f a c t , n e a r l y a l l such
f o r m u l a t i o n s i n v i t e f a l s e n e g a t i v e s a t a r a t e o f 50%.
The i n i t i a l NRC and IAEA f o r m u l a t i o n s d i f f e r e d a l s o w i t h
r e s p e c t t o t h e denominator. That i s , R e f e r e n c e 4 t r e a t s d e t e c t i o n
i n terms o f " s i g n a l s " , t h e denominator w h i c h c o n v e r t s from s i g n a l
u n i t s to concentration being only implied. I n c o n t r a s t , the
denominator f o r t h e LLD shows e x p l i c i t l y f a c t o r s f o r d e t e c t o r
E ( f f i c i e n c y ) , sample V(olume), c h e m i c a l Y ( i e l d ) , and r a d i o a c t i v e
D(ecay). The e s t i m a t e d s t a n d a r d d e v i a t i o n o f t h e b l a n k ( s ) i s t h e
B

CURRIE & PARR Perspectives on Detection Limits
U b
VO V> 00
to
H
W
^
rH
cd
C ^ rH
< m m
H cd
Q) S U 6
r-l
II I l i-H H
PU in PL
g
00

r-l

cd
II Pu
4-)
pi
CO b co m
CO VO

V)
H
> Pu
g Q
LL

fo

3 co
g
H \
4J
VI m vo
H Q cd pu
PL, w cd
s- to
C cd
er at]
VI e
3 y
^ VI
V4 pu
Re fe:rrenc
H
< b
4-> !
cd
4->
rH
\ rmH rmH m in 00 m b 3

< m CO r H rH| rH rH rH !

UIO

Pu
Csl
u !

VI
cd ) >
&
8 CO
4->
e

m 4J
H rH
PU ! 10
-m
CO I I
<t r-*
da
"
co oo r H CO CO cd
rH co CO r H r H J
m m in < "
cd co 1 1 I I rH I I 1
VI C PI Vi
b PS S Pu
I
11
CO CO cd (0 * I
<

SAMPLE MATRIX ELEMENTS OR CONCENTRATION SAMPLE CLASS REMARKS
CODE NUCLIDES OR ACTIVITY SIZE or
REFERENCED LEVEL (1 UNIT) SAMPLE"
BIOLOGICAL MATERIALS ( t e r r e s t r i a l )
A-ll Milk powder Trace elements: Ca v Environmental levels 25 g RM Non-certified information

Cl,Cu,Pe,Hg,K,Mg,Mn, values f o r : Ag, A l , As, Au,
Na,P,Rb,Zn B,Ba, Br, Cd, Cr, Ce, F, I, M
H
L l , Pb, Pt, Sb, Se, S i , Sn, M
Sr, T l , U, V
H
A-12 Animal bone Ra-226 5mBq/g (0.14 pCl/g) 80 g CRM

Sr-90 55mBq/g (1.48 pCi/g)
A-13 Freeze dried ani Trace elements : Br,Ca, Normal levels 25 g CRM Non-certlfled information >
mal blood Cu,Fe,K,Na,Rb,S,Se,Zn values f o r : Mg,Nl,P,Pb

>
A-14 Milk ponder Sr-90, Cs-137 Environmental levels 250 g CRM Non-certifled information
Na K, Ca
v value for Sr (total) 3
v-e >
Rye flour Trace elements: Br, Environmental levels 50 g RM Non-certifled information r
Ca,Cl,Cu,Fe,K,Mg,Mn, values f o r : Al,Au,Ba,Cd,Co,Cs,

P,Rb,Zn Mo,Na,S,Sb
V-9 Cotton cellulose Trace elements: Ba, Environmental levels 25 g CRM Non-certlfled information
e
Ca,CI,Cr,Cu,Hg,Mg,,
s
values f o r : Al,Br,Cd,Pe,Ga,
Mo,Na,Nl,Pb,Sr Hf,Ll,S,Sc,Se,Sm,Sn,Th,U,V S
W
V-10 Hay (powder) Trace elements: Ag, Environmental levels 50 g CRM Non-certifled information
As,BaBr,Ca Cd,Ce,Co
t values f o r : Al,Cs,Su,Hf,La,
Cr,Cu,F,Fe,Ga,Hg,I,K Mn,Sb,Se,Sm,Th
Mg,Mo,Na,Nl,P,Pb,Si
Sn,U,V,Zn
H-4 Animal muscle Trace elements: Br, Normal levels 20 g CRM

Ca,Cl,Cs,Cu,Fe,Hg,K, (2 v i a l s )
Mg,Mn,Na,Rb,Se,Zn
H-5 Animal bone ( i n c l u - Trace elements: Ba, Normal levels 30 g RM

ding mineral and Br,Ca,Cl,Fe,K,Mg,Na, (2 v i a l s )
organic components) P,Pb,Sr,Zn
H-8 Horse kidney Trace elements: Br, Normal levels 30 g RM Cd concentration similar to
Ca,Cd,Cl,Cu,Pe,Hg, human levels
K,Mg,Mn,Mo,Na,P,Rb, Non-certified information
Se,Zn values f o r : Co,Cs,S,Sr
H-9 Mixed human diet Multielement See remarks 30 g RM Representative of diet con-
sumed i n Finland.
Available from March 1986
* RM for Reference Material, CRM f o r C e r t i f i e d Reference Material
F i g u r e 1. B i o l o g i c a l R e f e r e n c e and C e r t i f i e d R e f e r e n c e M a t e r i a l s

a v a i l a b l e f r o m t h e I n t e r n a t i o n a l A t o m i c Energy Agency. (Reproduced
w i t h p e r m i s s i o n f r o m R e f . 2. C o p y r i g h t 1986, 1987 I n t e r n a t i o n a l
A t o m i c Energy Agency.)

Table I I . Detection Limits: P r a c t i c a l Needs
N u c l e a r R e g u l a t o r y Commission [NRC];
I n t e r n a t i o n a l Atomic Energy Agency [IAEA]
APPLICATION:
NRC Detection limits specified in the Technical

Specifications f o r n u c l e a r power r e a c t o r s f o r t h e
d e t e c t i o n o f e f f l u e n t and e n v i r o n m e n t a l r a d i o a c t i v i t y .
IAEA Detection l i m i t
toxic] i n bioenvironmental matrices. [Coordinated
R e s e a r c h Programs]
OBJECTIVE:
Method-independent definition and formulation; real

samples. ["cookbook" manual]
(a
INITIAL STATE > :
NRC -- LLD = 4 . 6 6 s /(EVY^D)

B
IAEA -- LOD - 3 s B
APPROACH:
NRC L i t e r a t u r e r e s e a r c h ; s i t e v i s i t s t o a s s e s s problems and

practices. [NRC r e g i o n a l o f f i c e s ; power r e a c t o r ;
contracting labs]
IAEA -- M u l t i d i s c i p l i n a r y consultants' m e e t i n g [IAEA HQ, Dec.

1985]
s represents
B t h e s t a n d a r d d e v i a t i o n o f t h e background o r b l a n k .
Other symbols i n d i c a t e : Efficiency, Volume, Y i e l d , and Decay
f a c t o r , resp.

9. CURRIE & PARR Perspectives on Detection Limits 177
s c a l i n g parameter i n b o t h e x p r e s s i o n s , though l i t t l e guidance i s

given f o r i t s determination. A l s o t h e NRC d e f i n i t i o n i s s l i g h t l y
ambiguous i n s t a t i n g t h a t s B i s "the standard d e v i a t i o n o f the
background...or... o f a b l a n k sample as a p p r o p r i a t e " ( 1 ) . Neither
approach explicitly t r e a t s p o s s i b l e non-normality, degrees o f
freedom, u n c e r t a i n t i e s i n t h e denominator [overall calibration
f a c t o r ] , s y s t e m a t i c e r r o r , o r t h e a p p l i c a t i o n t o multicomponent
systems which e x h i b i t i n t e r f e r e n c e and m a t r i x e f f e c t s and o f t e n
r e q u i r e more s o p h i s t i c a t e d c o m p u t a t i o n a l p r o c e d u r e s .
E f f e c t s o f t h e r e l a t i v e l y s i m p l i s t i c f o r m u l a t i o n s a r e seen, f o r
example, i n T a b l e I and F i g u r e s 2 and 3. I n the t a b l e , we see t h a t
o n l y two o f t h e r a d i o n u c l i d e s l i s t e d a r e commonly measured by
" s i m p l e c o u n t i n g " where a s i g n a l minus background o p e r a t i o n o b t a i n s .
F a r more commonly LLD's must be determined f o r multicomponent gamma
r a y s p e c t r o m e t r y , which may i n v o l v e r e l a t i v e l y c o m p l i c a t e d computa-
t i o n s and a t t e n d a n t assumptions r e g a r d i n g i n t e r f e r i n g components,
b a s e l i n e shapes, e t c
i s o l a t e d gamma r a y peaks
n o t always c o n s i s t e n t and erroneous d e t e c t i o n l i m i t s f o l l o w . One
p r a c t i c a l i l l u s t r a t i o n o f the p i t f a l l s attending the a p p l i c a t i o n of
the " s i m p l e c o u n t i n g " e x p r e s s i o n t o r e l a t i v e l y s i m p l e gamma r a y
s p e c t r o m e t r y has been d e s c r i b e d by R e i c h e l ( 5 ) . I n t h i s i n v e s t i g a -
t i o n , commercial spectrum a n a l y s i s s o f t w a r e was employed t o c a l c u -
l a t e t h e d e t e c t i o n l i m i t s o f Hg-203 (279 keV) and Cr-51 (320 keV) on
a Compton continuum from i n t e r f e r i n g Co-60. The recommended
d e t e c t i o n l i m i t o p t i o n which "has a p r o b a b i l i t y o f 95% o f [a peak]
b e i n g d e t e c t e d , " produced d e t e c t i o n l i m i t s o f 1557 and 1511 counts
f o r Hg-203 and Cr-51, r e s p e c t i v e l y . Y e t when peaks f o r these two
n u c l i d e s were added t o t h e continuum a t l e v e l s o f about 1580 and
1910 c o u n t s , r e s p e c t i v e l y , from a mixed s o u r c e , n e i t h e r was d e t e c t e d
by t h e s o f t w a r e , though b o t h were c l e a r l y v i s i b l e t o even an
u n t r a i n e d eye! See F i g u r e 2. The t h i r d peak a t 310 keV, d e r i v i n g
from t h e Co-60 Compton i n t e r a c t i o n s , was a l s o u n d e t e c t e d . The peaks
remained u n d e t e c t e d u n t i l t h e i r l e v e l s were a p p r o x i m a t e l y t w i c e t h e
presumed d e t e c t i o n l i m i t s , a t which p o i n t a good r e s u l t was o b t a i n e d
f o r Hg-203, b u t a poor one [ t o o s m a l l ] f o r Cr-51. This l a t t e r
r e s u l t no doubt f o l l o w e d from an erroneous b a s e l i n e assumption, even
though the program o p t i o n f o r "double peak e v a l u a t i o n " was chosen.
F i g u r e 3 i l l u s t r a t e s t h e problem f a c e d b y t h e IAEA i n t h e
b r o a d e r c o n t e x t o f t h e i r t r a c e element l a b o r a t o r y i n t e r c o m p a r i s o n
program. These d a t a show t h e r e p o r t e d r e s u l t s o f 16 l a b o r a t o r i e s
f o r measurements o f a r s e n i c i n t h e h o r s e k i d n e y i n t e r c o m p a r i s o n
sample (H-8), based on v a r i o u s v e r s i o n s o f atomic a b s o r p t i o n
spectrometry, o p t i c a l emission spectrometry, neutron activation
a n a l y s i s , and i n d u c e d X-ray e m i s s i o n a n a l y s i s . The o b j e c t i v e o f t h e
h o r s e k i d n e y i n t e r c o m p a r i s o n was t o a s s e s s (and r e f i n e ) a n a l y t i c a l
methods f o r t h e d e t e r m i n a t i o n o f e s s e n t i a l and t o x i c t r a c e elements
i n t h i s s u r r o g a t e f o r human k i d n e y ( 7 ) . K i d n e y , as the main t a r g e t
organ which accumulates t o x i c elements, was o f s p e c i a l i n t e r e s t w i t h
r e s p e c t t o cadmium. Horse k i d n e y , which c o n t a i n s s i m i l a r l e v e l s o f
cadmium t o the human k i d n e y c o r t e x , was s e l e c t e d f o r t h e development
and maintenance o f methods h a v i n g a demonstrated l e v e l o f q u a l i t y t o
a s s u r e r e l i a b l e b i o l o g i c a l m o n i t o r i n g o f t h i s element. Participants
were i n v i t e d t o a n a l y z e some 24 a d d i t i o n a l t r a c e elements, however,

F. Reichel (IAEA, 1985)
95% Confidence Level
250 300 350

Channel (keV)
F i g u r e 2. Non-detected, y e t q u i t e v i s i b l e gamma r a y peaks o f

Hg-203 and Cr-51 from IAEA p r a c t i c a l e x a m i n a t i o n o f s o f t w a r e
performance ( 5 ) . The continuum, shown dashed b e n e a t h t h e Hg-203
and Cr-51 peaks, i s due t o Co-60.
10*
105
9^
4
10
o. 10 3
X2 X2
9 S9
a. A5
99
10 2
AS ?9 9
10 1 99
10
? -I I I I I I I I L J I I L
40 30 36 3 52 27 25 38 54 73 35 61 6 12 10 29
Lab. Code No.
F i g u r e 3. I n t e r l a b o r a t o r y r e s u l t s f o r As, from t h e IAEA

i n t e r c o m p a r i s o n o f cadmium and o t h e r elements i n h o r s e k i d n e y
(H-8) ( 7 ) . F i l l e d c i r c l e s r e p r e s e n t q u a n t i t a t i v e r e s u l t s
( u n c e r t a i n t i e s n o t e x c e e d i n g c i r c l e d i a m e t e r ) ; open c i r c l e s
correspond t o reported detection l i m i t s .

most o f w h i c h are shown i n F i g u r e 1. R e s u l t s f o r cadmium, a t a

l e v e l o f about 190 mg/kg were q u i t e s a t i s f a c t o r y , b u t the much lower
concentrations o f a r s e n i c p r e s e n t e d d i f f i c u l t i e s . The 5 filled
c i r c l e s i n f i g u r e 3 r e p r e s e n t " q u a n t i t a t i v e " r e s u l t s -- i e , those
whose r e p o r t e d u n c e r t a i n t i e s do not exceed the d i a m e t e r o f the
circle. These are c l e a r l y i n c o n s i s t e n t , by more t h a n t h r e e o r d e r s
o f magnitude. Open c i r c l e s r e p r e s e n t " d e t e c t i o n l i m i t s " f o r non-
d e t e c t e d [ND] r e s u l t s ; t h e i r s p r e a d exceeds f o u r o r d e r s o f magni
tude. The c r o w n i n g i n c o n s i s t e n c y i s the f a c t t h a t a r s e n i c has been
"measured" by some l a b o r a t o r i e s a t l e v e l s c o n s i d e r a b l y i n excess o f
the d e t e c t i o n l i m i t s o f o t h e r s r e p o r t i n g ND -- c l e a r l y a l o g i c a l
impossibility.
To d e r i v e a c t u a l l e v e l s and e s t a b l i s h q u a l i t y measurement f o r
t r a c e elements i n f o o d s t u f f s , m a t e r i a l s such as m i l k powder (A-11)
and r e p r e s e n t a t i v e r e g i o n a l d i e t a r y b l e n d s (eg, USDIET-1) have been
provided (8.9). Consistency o f i n t e r l a b o r a t o r y r e s u l t s among
"independent" a n a l y t i c a
o f such m a t e r i a l s or th
n a t u r e o f the i n i t i a l measurement problem i s i l l u s t r a t e d by the
r e s u l t s o f the m i l k powder ( A - l l ) i n t e r c o m p a r i s o n , where the means,
r e l a t i v e ranges (max/min, d i m e n s i o n l e s s ) and medians o f 17 atomic
a b s o r p t i o n s p e c t r o s c o p y [AAS] and 7 r a d i o c h e m i c a l n e u t r o n a c t i v a t i o n
a n a l y s i s [RNAA] l a b o r a t o r y r e s u l t s f o r manganese were as f o l l o w s .
mean median
(mg/kg) max/min (mg/kg) (95% CI)
[AAS] 3.99 279 0.67 (0.45 - 1.27)

[RNAA] 0.295 4.8 0.26 (0.12 - 0.58)
I t i s i n t e r e s t i n g t o note t h a t the r a t i o o f means (AAS/RNAA) i s

13.5, whereas the r a t i o o f medians i s b u t 2.6. Reliability is
i n c r e a s e d t h r o u g h the use o f the median as a r o b u s t statistic,
e s p e c i a l l y when the number o f r e p l i c a t e s (here, l a b o r a t o r i e s ) i s
r e l a t i v e l y large. (When n>8, the 95% CI o f the median does not
d i r e c t l y depend upon the extremes, so some a d d i t i o n a l , a u t o m a t i c
o u t l i e r p r o t e c t i o n i s afforded.) I t i s n o t e w o r t h y , however, t h a t a
l a r g e number o f r e p l i c a t e s does not guarantee q u a l i t y ! The r e l a
t i v e l y few RNAA measurements y i e l d a more r e l i a b l e r e s u l t t h a n would
a v e r y e x t e n s i v e s e t o f AAS measurements ( r e p r e s e n t a t i v e o f t h i s
p a r t i c u l a r intercomparison). The b a s i c i s s u e i s one o f contamina
t i o n and i n t e r f e r e n c e , as i m p l i e d f o r example by the skewness
r e p r e s e n t e d i n the r a t i o o f mean t o median. T h i s "blank" i s s u e i s
b r o u g h t out a t t h i s p o i n t i n our d i s c u s s i o n because i t i s p r o b a b l y
the s i n g l e most i m p o r t a n t l i m i t a t i o n t o r e l i a b l e measurement and
r e l i a b l e d e t e c t i o n f o r l o w - l e v e l , multicomponent measurements i n
complex m a t r i c e s . As such, i t has been a major c o n s i d e r a t i o n i n
a d d r e s s i n g the q u e s t i o n o f " r e a l sample" d e t e c t i o n l i m i t s f o r b o t h
r a d i o l o g i c a l and t r a c e element measurements i n bioenvironmental
samples.
Experts and Concerned Groups. The foregoing observations

i n v o l v i n g d i v e r s e f o r m u l a t i o n s and i n t e r p r e t a t i o n s f o r r a d i o l o g i c a l
d e t e c t i o n l i m i t s , and d i v e r s e and d i s c r e p a n t i n t e r l a b o r a t o r y r e s u l t s

f o r low l e v e l t r a c e element measurements - - l e d the NRC and the

IAEA, r e s p e c t i v e l y , t o q u e s t i o n the v a l i d i t y or a t l e a s t the
u n i v e r s a l i t y o f d e t e c t i o n l i m i t f o r m u l a t i o n s as t h e n a p p l i e d . In
o r d e r t o address the q u e s t i o n i n an e f f i c i e n t manner the two
o r g a n i z a t i o n s employed problem s o l v i n g approaches o f demonstrated
e f f e c t i v e n e s s : v i z . , (a) s i t e v i s i t s and f i e l d i n t e r v i e w s w i t h each
o f the a f f e c t e d s e c t o r s [NRC], and (b) an i n t e n s i v e workshop
i n v o l v i n g e x p e r t s r e p r e s e n t i n g each o f the r e s p e c t i v e a n a l y t i c a l
disciplines [IAEA, ( 1 0 ) ] . The importance o f t h e s e approaches
d e s e r v e s emphasis, f o r i t t r a n s c e n d s the p a r t i c u l a r s c i e n t i f i c
problem under c o n s i d e r a t i o n . I n the case o f the NRC, f o r example,
the v a r y i n g needs and p e r c e p t i o n s o f the u t i l i t i e s and their
o p e r a t o r s , the c o n t r a c t ( r a d i o l o g i c a l ) l a b o r a t o r i e s , the c r o s s - c h e c k
( c o n t r o l ) l a b o r a t o r y , the i n s p e c t o r s , and NRC h e a d q u a r t e r s p e r s o n n e l
c o u l d not o t h e r w i s e have been d i s c e r n e d . As a r e s u l t i t was
p o s s i b l e t o i d e n t i f y and t o some e x t e n t s o r t out measurement from
p o l i c y i s s u e s , and attemp
l i m i t c o n c e p t s and t e r m i n o l o g
i n t e r p r e t a t i o n . A summary o f the f i n d i n g s o f the NRC s i t e v i s i t s i s
given i n Table I I I .
Table I I I . NRC Findings
* LLD Manual Needed - - f o r d i v e r s e backgrounds
* Wide r a n g i n g nomenclature and formulations

[LLD, MDA, ...]
* Socio-political-economic issues
[ b i a s e d r e p o r t i n g ( p u b l i c p e r c e p t i o n s ) , LLD r e q u i r e m e n t s for
minor components when h i g h i n t e r f e r e n c e l e v e l s , . . . ]
* Detection decisions: f a l s e negatives ranged from 5% t o 50%
* Blank: ambiguity in initial draft document, b l a n k v a r i a t i o n s
o f t e n e x c l u d e d from LLD
* Non-counting e r r o r s g e n e r a l l y ignored, e s p e c i a l l y sampling
* Simple c o u n t i n g f o r m u l a t i o n o n l y [ s i g n a l - b l a n k ] ; i n a p p l i c a b l e to
many cases o f n u c l e a r s p e c t r o m e t r y
* Gamma s p e c t r o m e t r y : multiple detection decisions; occasionally
h i d d e n , c h a n g i n g a l g o r i t h m s ; erroneous parameters
* Appropriate [ l o w - l e v e l ] and double b l i n d QA samples needed
The c a l l i n g together o f complementary or m u l t i d i s c i p l i n a r y

teams o f e x p e r t s i s a t r a d i t i o n i n a number o f i n t e r n a t i o n a l
o r g a n i z a t i o n s , such as the IAEA. S i n c e the IAEA o b j e c t i v e w i t h
r e s p e c t t o d e t e c t i o n l i m i t s was e x p l i c i t y "method independent", such
an approach was mandatory. That i s , e x p e r t knowledge c o n c e r n i n g the
n a t u r e and s o u r c e s o f e r r o r , such as the b l a n k , m a t r i x e f f e c t s , and
d e t e c t o r c h a r a c t e r i s t i c s f o r the methods o f i n t e r e s t (NAA, AAS,

ICP,...) was e s s e n t i a l so t h a t t h e approach t o d e t e c t i o n l i m i t s

would a t l e a s t be b r o a d enough t o encompass t h e c h a r a c t e r i s t i c s o f
t h e s e p a r t i c u l a r methods. A t h e o r e t i c a l f o r m u l a t i o n was n o t enough;
a common, p r a c t i c a l approach t o d e t e c t i o n , s u i t a b l e f o r t h e s e
s e v e r a l methods o f t r a c e a n a l y s i s was c a l l e d f o r . T a b l e IV g i v e s
the c o m p o s i t i o n and o b j e c t i v e s o f t h e IAEA t a s k f o r c e .
Outcomes
The b a s i c c o n c e p t s o f d e t e c t i o n d e c i s i o n s and d e t e c t i o n l i m i t s ,
based upon h y p o t h e s i s t e s t i n g as o u t l i n e d e a r l i e r i n t h i s volume,
n e c e s s a r i l y s e r v e d as t h e f o u n d a t i o n f o r t h e NRC and IAEA programs.
For t h e p r o b a b i l i s t i c p a r t o f t h e problem, f a l s e p o s i t i v e s and f a l s e
n e g a t i v e s were t a k e n e q u a l t o 0.05 i n each case. T h i s means t h a t
f o r normal random e r r o r s , t h e d e t e c t i o n d e c i s i o n i s made once t h e
o b s e r v e d s i g n a l exceed 1.645 time t h s t a n d a r d f th n u l l
s i g n a l [standard e r r o r
larger f o r paired observations]
freedom i s s m a l l , o f c o u r s e , t h e -statistic i s r e p l a c e d by
Student's-t. F o r s i m p l e measurements t h e d e t e c t i o n l i m i t i s j u s t
twice t h i s c r i t i c a l l e v e l . In addition to this r e l a t i v e l y straight
f o r w a r d n o r m a l , " s t a t i s t i c a l " t r e a t m e n t o f d e t e c t i o n , however, i t
was n e c e s s a r y t o pay a t t e n t i o n t o t h e s p e c i a l c h a r a c t e r i s t i c s o f t h e
r e a l measurement p r o c e s s e s i n v o l v i n g t h e d e t e c t i o n o f low l e v e l s o f
mixed r a d i o n u c l i d e s , and t r a c e c o n c e n t r a t i o n s o f m u l t i p l e elements
i n complex m a t r i x m a t e r i a l s . T h i s m a t t e r , w h i c h was an e x p l i c i t com
ponent o f b o t h p r o j e c t s , cannot be overemphasized. I f one c a l c u
l a t e s s t a t i s t i c a l d e t e c t i o n l i m i t s f o r i n t e r f e r e n c e - f r e e measure
ments i n pure s o l u t i o n s , e s t i m a t e d d e t e c t i o n l i m i t s w i l l be t o o low,
o f t e n by o r d e r s o f magnitude. S i m i l a r l y , d e t e c t i o n l i m i t s w h i c h a r e
p r e d i c t e d f o r an a l t e r a t i o n o r o p t i m i z a t i o n o f a measurement p r o c e s s
w i l l be u n r e a l i s t i c a l l y o p t i m i s t i c i f one a p p l i e s j u s t a s i m p l e
s t a t i s t i c a l f o r m u l a t o c a l c u l a t e t h e change i n d e t e c t i o n l i m i t , eg,
w i t h i n c r e a s e d r e p l i c a t i o n o r i n c r e a s e d c o u n t i n g time. Unless t e s t s
are made w i t h known o r common m a t e r i a l s comparable i n c o m p o s i t i o n t o
the samples o f i n t e r e s t , these erroneous e s t i m a t e s w i l l n o t be
exposed.
NRC S p e c i a l T o p i c s . The e s t i m a t i o n o f p r a c t i c a l , r e a l sample

d e t e c t i o n l i m i t s r e q u i r e s a t t e n t i o n t o a l l o f the sources o f e r r o r
t h a t must be f a c e d i n d e r i v i n g c o n f i d e n c e , o r more c o r r e c t l y ,
u n c e r t a i n t y i n t e r v a l s (11). I n t h e NRC s t u d y , s p e c i a l a t t e n t i o n was
g i v e n t o t h e f o l l o w i n g : a) bounds f o r uncompensated s y s t e m a t i c e r r o r
i n t h e background and c a l i b r a t i o n f a c t o r , b) n o n - P o i s s o n random
e r r o r ( t h a t w h i c h exceeds " c o u n t i n g s t a t i s t i c s " ) , c ) d e c o n v o l u t i o n
and model e r r o r connected w i t h a l p h a o r gamma-ray spectrum a n a l y s i s
o f multicomponent m i x t u r e s , and d) s p e c i a l l i m i t a t i o n s from t h e non-
normality o f the Poisson distribution f o r extreme low-level
c o u n t i n g , such as o c c u r s w i t h m o n i t o r i n g o f a c t i n i d e s o r o t h e r a l p h a
emitters. Each o f these f a c t o r s r e q u i r e d extension o f the
"simple-counting" expression o r i g i n a l l y found i n t h e d r a f t NRC
Technical S p e c i f i c a t i o n s [see t a b l e I I ] . A full discussion,
supplemented w i t h a p p r o p r i a t e f o r m u l a s , r e f e r e n c e s , and examples may
be found i n Ref. 12; and one o f t h e more b r o a d l y a p p l i c a b l e r e s u l t s
w i l l be t r e a t e d h e r e . I n t a b l e V we g i v e an e x t e n s i o n o f the

T a b l e IV. C o n s u l t a n t s ' M e e t i n g on L i m i t o f D e t e c t i o n
Dates and P l a c e
2-4 December 1985; IAEA H e a d q u a r t e r s , Vienna
Participants
Austria: W. Wegscheider
Belgium: J.P. Op de Beeck

R. Van G r i e k e n
F.R.G. M. S t o e p p l e r
U.S.A. L.A. Curri
IAEA R.M. P a r r ( S c i e n t i f i c Secretary)

F. R e i c h e l
R. Schelenz
H. V e r a - R u i z
METHODS
Topics A c t i v a t i o n A n a l y s i s and gamma-ray spectrometry

( J . P . Op de Beeck; F. R e i c h e l )
XRF and PIXE (R. Van G r i e k e n )
AAS (M. S t o e p p l e r )
ICP-AES (W. Wegscheider)
Voltammetry and o t h e r methods (M. S t o e p p l e r )
GOALS
General d i s c u s s i o n
Method-independent definition of the limit of

d e t e c t i o n (LoD)
P r a c t i c a l d e t e r m i n a t i o n and use o f t h e LoD
Special problems associated with individual

a n a l y t i c a l methods
Decision making and reporting of data below or

n e a r t h e LoD
P r e p a r a t i o n o f meeting r e p o r t

T a b l e V. D e s i g n and O p t i m i z a t i o n
LLD (LJJ ) V a r i a t i o n s w i t h background (B) and C o u n t i n g Time ( t ) ( a )
tr tr
< 1 Lp t ~ 1
Ln = c o n s t
h
1 < < 1000 LQ t ^ Lu t
> 1000 Lp - c o n s t ' LQ t
< a )
U n i t s of are counts. R and t r e p r e s e n t background r a t e and
B
sample c o u n t i n g t i m e , r e s p . , and i s a pure number determined by-

r e l a t i v e sample and background c o u n t i n g t i m e s . r r e p r e s e n t s the
mean l i f e ( t ^ / L n 2 ) . S p e c i a l l i m i t a t i o n s d e r i v e from the r e l a t i v e
systematic uncertaint
o f the P o i s s o n d i s t r i b u t i o
s y s t e m a t i c u n c e r t a i n t y i n the b l a n k [ ] .
p
[4A1 [ ^N] [ ]

2.71 + 3.29 jR^t + 0.10 R B t>

-()( [l-eÂ] >
\2.22 YEV/ \
s i m p l i f i e d e x p r e s s i o n , w h i c h now t a k e s i n t o account bounds f o r

p o s s i b l e s y s t e m a t i c background and c a l i b r a t i o n e r r o r , as w e l l as
moderate d e v i a t i o n o f P o i s s o n s t a t i s t i c s from n o r m a l i t y . The
symbols have the f o l l o w i n g meaning: Lp i s the d e t e c t i o n l i m i t i n
u n i t s of "concentration" i e , specific a c t i v i t y (pCi/L), following
R e f e r e n c e 1; , , V, t , and r r e p r e s e n t r a d i o c h e m i c a l y i e l d ,
c o u n t i n g e f f i c i e n c y , sample volume, c o u n t i n g t i m e , and mean (decay)
l i f e , r e s p e c t i v e l y ; R t - r e p r e s e n t s the number o f background
B
c o u n t s ; i s a pure number, r a n g i n g from 1 t o 2, w h i c h t a k e s i n t o

account the e f f e c t o f sample vs background c o u n t i n g times o r ( f o r
simple spectrometry) counting channels. Moderate d e v i a t i o n s o f
P o i s s o n s t a t i s t i c s from n o r m a l i t y a r e c o v e r e d by the term 2.71; and
=
the terms I * A N C
p r o v i d e f o r r e l a t i v e and a b s o l u t e

s y s t e m a t i c e r r o r bounds i n the o v e r a l l c a l i b r a t i o n f a c t o r [=YEV] and

background, r e s p e c t i v e l y .
The s t a n d a r d d e v i a t i o n o f the n u l l s i g n a l i n t h i s e x p r e s s i o n i s
g i v e n i n terms o f c o u n t i n g s t a t i s t i c s ; i f P o i s s o n s t a t i s t i c s a r e n o t
l i k e l y t o account f o r most o f the random c o u n t i n g e r r o r , t h e n i t
would be prudent t o deduce from a moderate number o f r e p l i c a t e s

-- i e , r e p l a c e the second term i n the numerator o f the second f a c t o r

by 2t's 77, where t ' i s S t u d e n t ' s - t and s i s the e s t i m a t e d s t a n d a r d
B B
d e v i a t i o n f o r the b l a n k ( c o u n t s ) . Bounds f o r s y s t e m a t i c e r r o r
s h o u l d be based on sound e x p e r i e n c e o r a n a l y s i s o f the measurement
p r o c e s s ; d e f a u l t v a l u e s t h a t r e f l e c t much l o w - l e v e l r a d i o n u c l i d e
measurement e x p e r i e n c e a r e s e t a t 1% [ b a s e l i n e ] , 5% [ b l a n k ] , and 10%
[calibration], respectively. P o i s s o n d e v i a t i o n s from n o r m a l i t y are
a d e q u a t e l y accounted f o r by t h i s e x p r e s s i o n down t o - 5 background

c o u n t s ; below t h i s , t a b u l a t e d v a l u e s f o r the P o i s s o n d i s t r i b u t i o n
s h o u l d be used [ ( 1 2 ) , pp 8 4 f f ] .
The importance o f the s y s t e m a t i c e r r o r bounds may be a p p r e c i a t e d
as the c o u n t i n g t i m e , sample s i z e o r number o f r e p l i c a t e s i s
increased. W i t h o u t the added term f o r b l a n k s y s t e m a t i c e r r o r ( l a s t
term i n the numerator o f the second f a c t o r ) , the d e t e c t i o n l i m i t
would decrease i n d e f i n i t e l y , f o r a l o n g - l i v e d n u c l i d e , s i m p l y by
e x t e n d i n g the c o u n t i n g time. A t the c o s t o f l o n g e r c o u n t i n g , a
f a l s e sense o f s e c u r i t y would r e s u l t . By way o f example, the
combined e f f e c t s o f a f i n i t e h a l f - l i f e and r e a l i s t i c s y s t e m a t i c
e r r o r bounds have been shown t o r e s t r i c t the improvement i n d e t e c
t i o n l i m i t f o r a s p e c i f i c 1-131 measurement t o a r e d u c t i o n o f about
25% even though the c o u n t i n g time i n c r e a s e d by a f a c t o r o f c a 100
(from 200 min t o the near o p t i m a l 2 weeks) [12, p. 137]. Table V
i l l u s t r a t e s the impact o f s y s t e m a t i c e r r o r , non-normal random e r r o r
[ P o i s s o n ] and decay c o n s t a n t [mean l i f e , r ] on c o u n t i n g - t i m e i n d u c e d
v a r i a t i o n s of radionuclid
v a r i a t i o n s can range w i d e l y
l i f e and background l e v e l .
B e f o r e l e a v i n g the t o p i c o f s y s t e m a t i c e r r o r bounds, two p o i n t s
s h o u l d be made. F i r s t , as i s perhaps o b v i o u s , the p r o b a b i l i s t i c
meaning o f f a l s e p o s i t i v e s and f a l s e n e g a t i v e s i s n e c e s s a r i l y
altered. These " e r r o r s " or r i s k s are now i n e q u a l i t i e s ["no g r e a t e r
t h a n . . . " ] , and t h e i r v a l i d i t y r e s t s g r e a t l y on t h a t o f the s y s
t e m a t i c e r r o r bounds, j u s t as i n the case o f u n c e r t a i n t y i n t e r v a l s
for high l e v e l s i g n a l s . Second, e s t i m a t i o n o f n o n - P o i s s o n random
e r r o r and s y s t e m a t i c e r r o r e m p i r i c a l l y , by comparison and r e p l i c a
t i o n i s n o t an easy t a s k . One can show t h a t a t l e a s t 15 and 47
r e p l i c a t e s , r e s p e c t i v e l y , are n e c e s s a r y j u s t t o d e t e c t s y s t e m a t i c
and excess random e r r o r components e q u i v a l e n t t o the (Poisson)
s t a n d a r d d e v i a t i o n [ ( 1 2 ) , 25f; ( 1 3 ) ] .
Most o f the e f f l u e n t and e n v i r o n m e n t a l r a d i o a c t i v i t y measurements
are made u s i n g gamma-ray s p e c t r o m e t r y . T h i s i s a f a r more c o s t
e f f e c t i v e approach than r a d i o c h e m i c a l a n a l y s i s ; the i n s t r u m e n t a l
measurement can be r e a d i l y automated, and d e t e c t i o n d e c i s i o n s can be
made more or l e s s s i m u l t a n e o u s l y f o r many r a d i o n u c l i d e s . The
v a l i d i t y o f those d e c i s i o n s , and o f the c o r r e s p o n d i n g d e t e c t i o n
l i m i t s , however, r e q u i r e s e i t h e r t h a t the peaks be i s o l a t e d and l i e
on a l i n e a r b a s e l i n e , or t h a t a d e t e c t i o n l i m i t model be employed
w h i c h i s more complex than t h a t used f o r " s i m p l e " c o u n t i n g .
B a s e l i n e o r i n t e r f e r e n c e model u n c e r t a i n t i e s s h o u l d be i n c l u d e d , and
an i t e r a t i v e s o l u t i o n i s r e q u i r e d t o e s t i m a t e the d e t e c t i o n l i m i t
when spectrum d e c o n v o l u t i o n i s i n v o l v e d . D e t a i l s a r e beyond the
scope o f t h i s c h a p t e r , b u t a r e l a t i v e l y s i m p l e l i m i t i n g e s t i m a t e can
be d e r i v e d by t r e a t i n g the e s t i m a t e d s t a n d a r d e r r o r f o r a low l e v e l
r a d i o n u c l i d e peak o f i n t e r e s t as though i t were the n u l l s t a n d a r d
error, a Q [12, p. 8 1 ] .
Multicomponent gamma-ray s p e c t r o m e t r y i s subject to s e v e r a l
a d d i t i o n a l d e t e c t i o n l i m i t p i t f a l l s w h i c h w i l l s i m p l y be n o t e d h e r e .
I f the a l g o r i t h m changes i n p a s s i n g from peak d e t e c t i o n t o peak
e s t i m a t i o n , an i n v a l i d d e t e c t i o n l i m i t w i l l be g i v e n , u n l e s s a l g o
r i t h m s w i t c h i n g i s p r o p e r l y taken i n t o c o n s i d e r a t i o n . Incorrect
f a l s e p o s i t i v e and/or f a l s e n e g a t i v e e r r o r s w i l l r e s u l t from non
l i n e a r peak s e a r c h r o u t i n e s , erroneous p e a k / b a s e l i n e models, and
s u b t l e b u t o f t - h i d d e n d e v i a t i o n s from the b a s i c h y p o t h e s i s t e s t i n g

principles. Since algorithms are not always a v a i l a b l e f o r

examination, and since even the raw s p e c t r a l d a t a may be
n o n - r e t r i e v a b l e , these a u t o m a t i c s p e c t r a l e v a l u a t i o n t e c h n i q u e s can
sometimes generate m i s l e a d i n g d e t e c t i o n l i m i t e s t i m a t e s . (See the
f o r e g o i n g d i s c u s s i o n o f F i g . 2.) The impact o f m u l t i p l e d e t e c t i o n
d e c i s i o n s on f a l s e p o s i t i v e s and n e g a t i v e s i s a n o t h e r i s s u e t h a t
must be f a c e d i n l a r g e bandwidth (many p o t e n t i a l component) systems
( 1 4 ) . For example, i f one were to s e a r c h a l a r g e s p e c t r a l r e g i o n
f o r , say, 100 peaks w h i c h are t r u l y absent, the p r o b a b i l i t y o f a
f a l s e p o s i t i v e would be p r a c t i c a l l y u n i t y (99.4%) u n l e s s the
c r i t i c a l level were a p p r o p r i a t e l y a d j u s t e d .
IAEA S p e c i a l T o p i c s . U n l i k e the NRC, the IAEA was concerned

p r i m a r i l y w i t h the a c t u a l d e t e c t i o n l i m i t s o b t a i n e d by program
p a r t i c i p a n t s , as p r a c t i c a l performance c h a r a c t e r i s t i c s o f t h e i r
methods as a p p l i e d t o comple
r e g u l a t o r y l i m i t (1^ ) d i
o p t i m i z a t i o n or r e d e s i g n sampl analyte
t r a c e a n a l y s i s methods were somewhat b r o a d e r t h a n those c o n s i d e r e d
by the NRC. That, p l u s the r e q u i r e m e n t f o r a "method independent"
approach, meant t h a t the IAEA had to g i v e somewhat more s c r u t i n y to
the f o i b l e s o f the s e v e r a l a n a l y t i c a l methods, e s p e c i a l l y w i t h
r e s p e c t to b l a n k s , l o s s e s , i n t e r f e r e n c e and m a t r i x e f f e c t s , and
detector c h a r a c t e r i s t i c s . Methods e x p e r t s , r e f e r r e d to above,
p r o v i d e d c r i t i c a l i n f o r m a t i o n on e x p l i c i t problems, such as uncer
t a i n t i e s i n b a s e l i n e or peak o v e r l a p models i n NAA and XRF methods
of a n a l y s i s . Such added s o u r c e s o f e r r o r mean, f o r example, t h a t
d e t e c t i o n l i m i t s d e r i v e d s t r i c t l y from spectrum f i t t i n g variance
w i l l be too s m a l l . Numerous a d d i t i o n a l m e t h o d - s p e c i f i c s o u r c e s o f
e r r o r , ranging from d i s s o l u t i o n and r e c o v e r y to matrix-induced
r a d i a t i o n s c a t t e r i n g were d e v e l o p e d by the e x p e r t s , and w i l l be
summarized i n the subsequent r e p o r t (10).
The b l a n k and c o n t a m i n a t i o n problem was h i g h l i g h t e d i n the IAEA
deliberations. I t seems l i k e l y t h a t t h i s i s the most s i g n i f i c a n t
cause o f the o r d e r s o f magnitude d i s c r e p a n c i e s i n d e t e c t i o n l i m i t s
r e p o r t e d i n IAEA i n t e r c o m p a r i s o n s . To d e a l e f f e c t i v e l y w i t h t h i s
problem, s e v e r a l recommendations were made. F i r s t , f u t u r e intercom
p a r i s o n p a r t i c i p a n t s w i l l be encouraged to measure and report
r e s u l t s f o r a minimum number o f b l a n k s , perhaps n i n e . Second,
d e t e c t i o n l i m i t s w i l l be based on p a i r e d comparisons o f sample
measurements w i t h e q u i v a l e n t b l a n k measurements, and the average o f
a t l e a s t t h r e e or f o u r (but p r e f e r a b l y n i n e o r more) such p a i r e d
comparisons w i l l be used ( w i t h S t u d e n t ' s - t ) f o r d e t e c t i o n d e c i s i o n s .
The p h i l o s o p h y u n d e r l y i n g t h i s approach i s t h a t : a) p a i r e d com
p a r i s o n s f o r c e symmetry i n the e s t i m a t e d n e t s i g n a l s , y i e l d i n g a
b e t t e r chance t o approach n o r m a l i t y t h r o u g h a v e r a g i n g [ c e n t r a l l i m i t
theorem], p l u s the p o s s i b i l i t y o f u s i n g the Gauss i n e q u a l i t y ( i n
c o n t r a s t to the w i d e r T s c h e b y s c h e f f i n e q u a l i t y ) f o r d i s t r i b u t i o n
f r e e l i m i t e s t i m a t i o n ; and b) w i t h n i n e or more o b s e r v a t i o n s , the
median and i t s 95% c o n f i d e n c e i n t e r v a l can be used f o r r o b u s t
e s t i m a t i o n w i t h l i t t l e i n f l u e n c e from p o s s i b l e o u t l i e r s . Finally,
the d e t e c t i o n l i m i t can be e s t i m a t e d as a p p r o x i m a t e l y t w i c e the
c r i t i c a l l e v e l or 2ta . When i s e s t i m a t e d as 's' v i a r e p l i c a t i o n ,
Q
and one wishes a c o n s e r v a t i v e upper l i m i t f o r L^ , t h i s can be formed

2
t h r o u g h the combined use of the Gauss inequality and the
distribution.
L e t us i l l u s t r a t e . Assume t h a t a s e t o f t e n p a i r e d o b s e r v a t i o n s
o f the t o t a l s i g n a l (T - S + B) and the b l a n k (B) f o r the a n a l y s i s
o f t r a c e l e v e l s o f Zn i n a n i m a l bone ( i n u n i t s o f mg/kg) were as
follows
B: 164.9 144.7 135.1 139.1 167.6 246.3 228.8 111.9 150.3 153.0
T: 203.9 277.7 288.7 202.9 164.2 227.4 241.3 262.2 238.4 250.3
The estimated means and standard errors are
Mean fclank - 164.2 SE - 13.3 (mg Zn/kg bone)

Mean S (T-B) - 71.5 SE - 20.0 (mg Zn/kg bone)
For 10 degrees o f freedom S t u d e n t ' s - t at the (1-sided) 5%

s i g n i f i c a n c e l e v e l i s 1.81
36.2, and one would c o n c l u d
g r e a t e r t h a n t h ^ b l a n k -- i e , Zn was d e t e c t e d . (Note t h a t s
Q was
t a k e n as SE o f S above, a v a l i d p r o c e d u r e , so l o n g as S i s s m a l l ,
i e , below the d e t e c t i o n l i m i t . A l t e r n a t i v e l y , i t c o u l d have been
t a k e n as 72>SE - 18.8, w i t h 9 degrees o f freedom.) The complemen
B
t a r y q u e s t i o n i s : "What i s the s m a l l e s t c o n c e n t r a t i o n o f Zn t h a t
c o u l d be d e t e c t e d w i t h 95% p r o b a b i l i t y , g i v e n the above c r i t i c a l
level?" The answer, namely the e s t i m a t e d d e t e c t i o n l i m i t f o r the
measurement p r o c e s s under c o n s i d e r a t i o n , i s a p p r o x i m a t e l y t w i c e the
c r i t i c a l l e v e l o r 72.4 mg/kg. Note t h a t 71.5 i s an experimental
outcome (which we t e s t e d f o r s t a t i s t i c a l s i g n i f i c a n c e ) , whereas 72.4
i s a measure o f the i n h e r e n t d e t e c t i o n c a p a b i l i t y o f the measurement
process. An upper l i m i t f o r Lp may be computed u s i n g the upper
2
l i m i t f o r /s. U s i n g w i t h 10 degrees o f freedom, we f i n d a v a l u e
o f 1.59 f o r t h i s r a t i o ( 1 - s i d e d , 5% s i g n i f i c a n c e l e v e l ) , hence an
upper l i m i t f o r Lp o f 115 mg Zn/kg bone. The t r u e d e t e c t i o n l i m i t
i s l i k e l y (95% chance) s m a l l e r t h a n t h i s v a l u e , b u t a more p r e c i s e
e s t i m a t e o f would be r e q u i r e d t o b e t t e r d e t e r m i n e i t . A s t i l l
more c o n s e r v a t i v e v a l u e , i n c o r p o r a t i n g the Gauss i n e q u a l i t y where
1
[ t + ( 1 . bjlj})' ]a Q r e p l a c e s 2to , would r a i s e t h i s upper l i m i t f o r
Q
the d e t e c t i o n l i m i t by an a d d i t i o n a l 8%, t o 124 mg/kg. The major

c o n s t r a i n t on p r e c i s e knowledge o f the d e t e c t i o n l i m i t i s thus
knowledge o f ; f o r n o r m a l l y d i s t r i b u t e d d a t a , w i t h s b a s e d on

r e p l i c a t i o n , the u n c e r t a i n t y range ( a t the 90% c o n f i d e n c e l e v e l ) f o r

/s f a l l s below a f a c t o r o f 2 once the number o f r e p l i c a t e s exceeds
a dozen.
Two points merit emphasis i n the above e x e r c i s e : a) The
A
s t a t i s t i c a l c o n f i d e n c e i n t e r v a l f o r the outcome Js b a s e d on S and

i t s SE ( u s i n g a 2 - s i d e d S t u d e n t ' s - t ) ; SE b u t n o t S i s u s e d a l s o f o r
the e s t i m a t i o n o f 1^ . b) The c o n f i d e n c e i n t e r v a l , and and Lp
(and i t s upper l i m i t ) are c o r r e c t f o r n o r m a l l y d i s t r i b u t e d random
errors. P a i r e d , comparisons and a moderate number o f r e p l i c a t e s
t e n d t o make t h e s e assumptions r e a s o n a b l y good; t h i s i s an i m p o r t a n t
p r e c a u t i o n , g i v e n the w i d e l y v a r y i n g b l a n k d i s t r i b u t i o n s o f such
d i f f i c u l t measurements. Perhaps the most i m p o r t a n t consequence o f
the p a i r e d c o m p a r i s o n i n d u c e d symmetry, i s t h a t the e x p e c t e d v a l u e
A
f o r the n u l l s i g n a l [ - B'] w i l l be z e r o -- i e , u n b i a s e d .
S y s t e m a t i c e r r o r bounds, some deeper implications of paired

comparisons, and d i s t r i b u t i o n f r e e t e c h n i q u e s a r e t r e a t e d i n t h e
extended d i s c u s s i o n o f t h e above example ( 1 0 ) . C o n s i d e r a t i o n o f
r e a s o n a b l e s y s t e m a t i c e r r o r bounds f o r t h e b l a n k and o v e r a l l
c a l i b r a t i o n f a c t o r l e d t o a f i n a l d e t e c t i o n l i m i t [upper l i m i t f o r ]
o f about 140 mg/kg. Use o f t h e median as a more a s s u m p t i o n - f r e e
e s t i m a t o r y i e l d e d a 95% c o n f i d e n c e i n t e r v a l e x t e n d i n g from -3 ( e r g o ,
0) t o 150 mg/kg. (The n u m e r i c a l v a l u e s f o r t h i s s i m u l a t e d example
were i n s p i r e d b y a c t u a l IAEA i n t e r c o m p a r i s o n d a t a f o r Zn i n a n i m a l
bone [H-5], as r e p o r t e d i n T a b l e I o f Ref. 6.)
G e n e r a t i n g a u t h e n t i c b l a n k s f o r measurement i s another m a t t e r .
The c e n t r a l importance o f t h e b l a n k f o r r e l i a b l e a n a l y t e d e t e c t i o n ,
and t h e c o m p l e x i t y o f t h e b l a n k i n multicomponent t r a c e element
a n a l y s i s o f b i o e n v i r o n m e n t a l m a t r i c e s a r e such t h a t t h e IAEA gave
t h i s matter s p e c i a l a t t e n t i o n . A t h r e e f a c e t e d approach was
devised, comprising the " i d e a l blank," the " s i m u l a t i o n or surrogate
b l a n k , " and f i n a l l y " p r o p a g a t i o n o f t h e b l a n k . " F u l l discussion
i n c l u d i n g method s p e c i f i
a b r i e f exposition follows
which r e f l e c t s t h e sample and t h e measurement p r o c e s s i n e v e r y
r e s p e c t save one: t h e absence o f t h e a n a l y t e o f i n t e r e s t . For
r e l a t i v e l y s i m p l e m a t r i c e s and r e l a t i v e l y l o w l e v e l s o f i n t e r -
f e r e n c e , t h e i d e a l b l a n k may be approached. F a i l i n g t h i s , one must
d e v i s e s u r r o g a t e b l a n k s which c l o s e l y s i m u l a t e t h e r e a l sample.
T h i s p r o c e s s , as w e l l as t h e a l t e r n a t i v e b l a n k p r o p a g a t i o n t e c h -
n i q u e , r e q u i r e s t r u e e x p e r t i s e i n t h e r e l e v a n t a n a l y t i c a l s c i e n c e as
opposed t o s t a t i s t i c s . Remembering t h a t t h e " b l a n k e f f e c t " -- i e ,
t r u e a n a l y t e b l a n k t o g e t h e r w i t h u n r e s o l v e d interfrants -- may be
a s s o c i a t e d w i t h t h e sample, i t s e l f , d i s s o l u t i o n p r o c e d u r e s , r e a -
g e n t s , and i n s t r u m e n t a l (and even s o f t w a r e ) a r t i f a c t s , we note t h a t
a good s u r r o g a t e b l a n k r e q u i r e s t h e i n t r o d u c t i o n o f a n a l y t e o r
interfrant a t t h e same s t a g e s o f t h e measurement p r o c e s s and i n t h e
same amounts as o c c u r w i t h t h e a c t u a l sample. Spectral inter-
f e r e n c e s , m a t r i x e f f e c t s , r e c o v e r i e s , s e n s i t i v i t i e s , and r e a g e n t
q u a n t i t i e s and sample s i z e s s h o u l d be s u f f i c i e n t l y s i m i l a r . The
a l l o w a b l e degree o f d e p a r t u r e , a g a i n i s a sample-method s p e c i f i c
s c i e n t i f i c i s s u e which must be determined b y t h e r e s p e c t i v e e x p e r t s ,
perhaps w i t h t h e a i d o f e x p e r i m e n t a l ruggedness t e s t s .
An i n t e r e s t i n g i l l u s t r a t i o n o f t h e s u b t l e t y o f s u r r o g a t e b l a n k
p i t f a l l s i s found i n t h e a n a l y s i s o f t r a c e elements i n m i l k by
neutron a c t i v a t i o n a n a l y s i s . Knowledge o f t r a c e c o n s t i t u e n t s o f
m i l k , e s p e c i a l l y human m i l k , i s o f c o n s i d e r a b l e importance i n IAEA
c o o p e r a t i v e r e s e a r c h programs i n v o l v i n g g l o b a l t r e n d s i n human
h e a l t h and n u t r i t i o n ( 3 . 8 ) . L a c k i n g s u f f i c i e n t i n f o r m a t i o n r e g a r d -
i n g 1) b a s e l i n e i n t e r f e r e n c e i n NAA u s i n g gamma r a y s p e c t r o m e t r y ,
and 2) t r a c e element c o m p o s i t i o n o f presumably s i m i l a r b i o l o g i c a l
m a t e r i a l s , one might s e l e c t cows' m i l k as a p o t e n t i a l s u r r o g a t e .
Such a c h o i c e would be m i s l e a d i n g f o r a number o f a c t i v a t e d gamma
e m i t t e r s , however, because cows' m i l k c o n t a i n s phosphorus a t con-
c e n t r a t i o n s which exceed those i n human m i l k b y about a f a c t o r o f
10, and t h e b r e m s s t r a h l u n g from t h e h i g h energy b e t a e m i t t i n g
n e u t r o n a c t i v a t i o n p r o d u c t P-32 would cause a much i n c r e a s e d i n s t r u -
mental b a s e l i n e c o n t r i b u t i o n which would n o t be r e p r e s e n t a t i v e o f
the t r u e , human m i l k b l a n k ( 1 5 ) . F o r o t h e r methods, such as AAS o r
XRF, t h i s problem would n o t a r i s e .

Space does n o t p e r m i t a d e t a i l e d d i s c u s s i o n o f t h e p r o p a g a t i o n
approach t o t h e b l a n k and i t s v a r i a b i l i t y , b u t i n essence i t
c o n s i s t s o f c o n s i d e r i n g each s e q u e n t i a l s t e p o f t h e measurement
p r o c e s s , t o g e t h e r w i t h t h e i n t r o d u c t i o n and p r o p a g a t i o n o f a n a l y t e
b l a n k and interfrants through each s t e p , t a k i n g i n t o account t h e
r e s p e c t i v e r e c o v e r i e s . I n t h e f i n a l , i n s t r u m e n t a l measurement s t e p
p e r t u r b a t i o n s o f t h e response (shape and s e n s i t i v i t y ) f o r a n a l y t e ,
interfrants, and m a t r i x a b s o r p t i o n and s c a t t e r i n g must a l l be
considered. P r o p a g a t i o n o f components o f t h e b l a n k thus r e p r e s e n t s
an e x c e l l e n t independent approach t o d e v i s i n g a s u r r o g a t e b l a n k .
I t s s u c c e s s , however, depends upon t h e e s t i m a t i o n o f a p p r o p r i a t e
r e c o v e r y and s e n s i t i v i t y f a c t o r s f o r t h e t h r e e t r o u b l e makers:
a d v e n t i t i o u s a n a l y t e , interfrants, and m a t r i x e f f e c t s . Excellent
e x p e r i m e n t a l t e c h n i q u e s w h i c h may h e l p i n t h e t a s k i n c l u d e m u l t i p l e
s t a n d a r d a d d i t i o n s , i s o t o p e d i l u t i o n , and m u l t i p l e sample a d d i t i o n s
(16) .
R e p o r t i n g o f Low-Level Data
R e p o r t i n g and q u a l i t y i s s u e s t r a n s c e n d t h e s p e c i f i c NRC and IAEA

programs, and they a r e t r e a t e d i n some d e t a i l elsewhere i n t h i s
volume. L o w - l e v e l d a t a a r e f a r more s u b j e c t t o i n f o r m a t i o n l o s s and
b i a s than d a t a c o r r e s p o n d i n g t o l a r g e s i g n a l s . T h i s f o l l o w s because
o f a tendency t o r e p o r t o b s e r v a t i o n s w h i c h a r e n o t s t a t i s t i c a l l y
s i g n i f i c a n t as z e r o , o r " t r a c e " , o r as v a r i o u s types o f upper
l i m i t s . A v e r a g i n g o r combining o r even comparing s e t s o f r e s u l t s so
reported i s e i t h e r impossible or not p o s s i b l e without b i a s . Worse
s t i l l , t h e same r e s u l t s c o u l d be r e p o r t e d a c c o r d i n g t o d i f f e r e n t
p r e s c r i p t i o n s , l e a d i n g a t t h e v e r y l e a s t t o m i s c o n c e p t i o n s by t h e
lay public. The recommendation g i v e n b o t h t h e NRC and t h e IAEA was
t h a t observed v a l u e s and t h e i r u n c e r t a i n t i e s be r e p o r t e d , t o g e t h e r
with appropriate detection l i m i t i n f o r m a t i o n , even when t h e
detection d e c i s i o n i s negative.
Q u a l i t y c o n t r o l o r c r o s s check samples a t low l e v e l s a r e
e s s e n t i a l f o r a s s u r i n g q u a l i t y measurements i n these same con-
centration regions. I n t e r n a l , known c o n t r o l s i n s i m i l a r m a t r i c e s
and h a v i n g s i m i l a r i n t e r f e r e n c e s s h o u l d be t h e f i r s t s t e p toward
attaining control. C e r t i f i e d , n a t u r a l m a t r i x m a t e r i a l s are the
b e s t , f o r one c a n then t e s t n o t j u s t c o n s i s t e n c y o r i n t e r l a b o r a t o r y
c o m p a r a b i l i t y , but a l s o accuracy. Use o f e x t e r n a l , b l i n d c o n t r o l
samples, perhaps i n c l u d i n g l a b o r a t o r y i n t e r c o m p a r i s o n s , may be t h e
o n l y way t o r e l i a b l y a s s e s s performance under r o u t i n e c o n d i t i o n s
(17) . S i m i l a r l y , t e s t d a t a s e t s can be most i m p o r t a n t f o r a l g o r i t h -
mic q u a l i t y assurance o f low l e v e l multicomponent spectrometry
(18.19) . I t was f e l t t h a t t h e o b j e c t i v e s o f b o t h t h e NRC and IAEA
programs c o u l d be b e t t e r met by t a k i n g c a r e f u l measures t o a s s u r e
t r u l y b l i n d c o n t r o l samples t h a t would be f a i t h f u l s u r r o g a t e s w i t h
r e s p e c t t o a n a l y t e c o n c e n t r a t i o n and m a t r i x c o m p o s i t i o n f o r t h e
samples and a n a l y t e s o f concern. D i s t r i b u t i o n o f b l i n d blanks c o u l d
f u r t h e r assure v a l i d claims concerning d e t e c t i o n l i m i t s .
Conclusion
The NRC and IAEA programs share the common p r a c t i c a l g o a l o f s e e k i n g

t o e s t a b l i s h a c c u r a t e v a l u e s f o r t h e d e t e c t i o n l i m i t performance

c h a r a c t e r i s t i c f o r t h e i r r e s p e c t i v e measurement p r o c e s s e s f o r t r a c e
r a d i o n u c l i d e s and t r a c e elements, r e s p e c t i v e l y . The d r i v i n g f o r c e s
f o r e s t a b l i s h i n g m e a n i n g f u l d e t e c t i o n l i m i t s f o r t h e a c t u a l measure
ment p r o c e s s were, i n t h e f i r s t case f o r r a d i o l o g i c a l m o n i t o r i n g t o
prevent the r e l e a s e o f r a d i o a c t i v i t y l e v e l s o f environmental
c o n c e r n , and i n t h e second, t o i n c r e a s e o u r knowledge o f c o n c e n t r a
t i o n s and g e o g r a p h i c v a r i a t i o n s o f e s s e n t i a l and t o x i c t r a c e
elements i n b i o l o g i c a l and e n v i r o n m e n t a l samples.
D e t e c t i o n l i m i t f o r m u l a t i o n s w h i c h appeared s i m i l a r , b u t w h i c h
were f u n d a m e n t a l l y d i f f e r e n t , were i n i t i a l l y found. I n b o t h cases
the e x p r e s s i o n s were a l s o somewhat l i m i t e d i n a p p l i c a b i l i t y , i n t h a t
they d i d n o t e x p l i c i t l y t r e a t s y s t e m a t i c e r r o r s , o r those a s s o c i a t e d
w i t h many o f t h e " r e a l - l i f e " problems o f l o w - l e v e l measurement such
as o v e r l a p p i n g peaks, m a t r i x e f f e c t s , b l a n k i n t r o d u c t i o n a t v a r i o u s
s t a g e s o f measurement, i n s t r u m e n t a l a r t i f a c t s , o r non-normal random
errors. A s y n o p s i s o f t h e i s s u e s t o g e t h e r w i t h an i n d i c a t i o n o f
the c r u c i a l r o l e o f s c i e n t i f i
b e n e f i t s o f sound measuremen
d e r i v a t i o n o f meaningful d e t e c t i o n l i m i t s . I n F i g u r e 4, t o be
c o n t r a s t e d w i t h F i g u r e 3, we i l l u s t r a t e a "success s t o r y . " Here,
the improvement i n measurement q u a l i t y has made p o s s i b l e t h e
d i s c r i m i n a t i o n o f s i g n i f i c a n t geographical v a r i a t i o n s i n the trace
Table V I . Outcome and D e s i d e r a t a
NRC -- Method-independent o f o r m u l a t i o n ; a t t e n t i o n t o n o n - c o u n t i n g
Q
random and s y s t e m a t i c e r r o r . S p e c i a l t r e a t m e n t f o r few

counts
IAEA -- P a i r e d o b s e r v a t i o n s ; r e p o r t an adequate number o f " i d e a l "
b l a n k s ; use o f " t " , c e n t r a l l i m i t theorem, and /s ( l i m i t )
1. Sound, c o n c e p t u a l b a s i s [not ad hoc]
2. Method-independent t e r m i n o l o g y , f o r m u l a t i o n
3. Responsible r e p o r t i n g [low-level data]
4. Use o f t h e f u l l power o f s t a t i s t i c s
[hypothesis t e s t i n g , robust e s t i m a t i o n , e r r o r propagation...]
5. GOOD SCIENCE
Knowledge o f t h e Measurement P r o c e s s i n place of empiricism

["local" matrix c a l i b r a t i o n and l a r g e r e p l i c a t i o n s , a r e
costly]
E x p e r t knowledge r e : simulation blank, propagation of

components o f t h e b l a n k .
Special a t t e n t i o n to error bounds [ i n c l u d i n g m o d e l - e r r o r ] ,

v a l i d a t i o n , c o n t a m i n a t i o n , i n t e r f e r e n c e , and m a t r i x e f f e c t s
A n a l y t i c a l Q u a l i t y A s s u r a n c e v i a l o w - l e v e l C e r t i f i e d Reference
M a t e r i a l s , and S i m u l a t i o n T e s t Data [ a l g o r i t h m i c QA].

100
U.S. ( 0)
2
ce
UJ
.
(/)
<
ce

ce

s
0C
UJ

COUNTRY
F i g u r e 4. Manganese c o n c e n t r a t i o n s i n human m i l k from d i f f e r e n t

c o u n t r i e s , medians SD (median) ( 2 0 ) . Shaded r e g i o n superposed
r e p r e s e n t s t h e " i n n e r range" o f t h e A-11 m i l k powder
i n t e r c o m p a r i s o n (8.), f o r 27 r e s u l t s e x c l u d i n g one major o u t l i e r
and t h e two r e m a i n i n g extreme v a l u e s , t r a n s l a t e d b y median
m a t c h i n g . That i s , t h e median o f 21 "accepted" A - l l r e s u l t s , i n
u n i t s o f %/% o f m i l k powder, was t r a n s l a t e d t o match t h e median
o f t h e s i x r e g i o n a l l i q u i d human m i l k r e s u l t s , i n u n i t s o f /ig/L,
i n o r d e r t o compare r e l a t i v e ranges. D i f f e r e n c e s i n t h e
n u m e r i c a l v a l u e s o f t h e two medians a r e due i n l a r g e p a r t t o t h e
w a t e r c o n t e n t o f t h e l i q u i d m i l k . The range o f t h e 21
"accepted" A - l l r e s u l t s was s m a l l e r , e q u i v a l e n t t o a f a c t o r o f
6, b u t t h i s s t i l l would have been i n a d e q u a t e f o r d e t e c t i o n o f
the g e o g r a p h i c v a r i a t i o n s shown. The arrow a t 50 g/L
c o r r e s p o n d s t o t h e s u g g e s t e d l i m i t f o r Mn i n U.S. d r i n k i n g w a t e r
(21).

element c o m p o s i t i o n o f human m i l k ( 2 0 ) . Knowledge o f t h i s s o r t , as

s t a t e d e a r l i e r , c o n s t i t u t e s one o f the p r i m a r y g o a l s o f the IAEA
c o o r d i n a t e d r e s e a r c h programs. I t can l e a d t o b e t t e r u n d e r s t a n d i n g
o f n u t r i t i o n a l dficiences through c o r r e l a t i o n w i t h s t a t e s of
h e a l t h , and more i m p o r t a n t l y , can l e a d the way toward improved
g l o b a l h e a l t h through a p p r o p r i a t e d i e t a r y supplements d u r i n g the
c r u c i a l f i r s t y e a r s o f l i f e . The shaded r e g i o n i n F i g u r e 4 shows the
c o n n e c t i o n w i t h the e a r l i e r d e t e c t i o n and measurement c a p a b i l i t e s
f o r the same element (Mn) i n the IAEA A - l l m i l k powder i n t e r c o m -
p a r i s o n (8) . T h i s r e g i o n c o n s t i t u t e s the i n n e r range ( e x c l u d i n g a
major o u t l i e r p l u s the two r e m a i n i n g extremes) r e p r e s e n t i n g 27
l a b o r a t o r i e s , s c a l e d by median matching. That i s , the median o f 21
"accepted" r e s u l t s f o r Mn i n the A - l l M i l k Powder, 0.32 /ig/g, was
a l i g n e d w i t h the median o f the g e o g r a p h i c r e s u l t s f o r human m i l k , c a
7 /ig/L, i n o r d e r t o compare the r e l a t i v e ranges. C l e a r l y , the
i m p o r t a n t g e o g r a p h i c v a r i a t i o n s would n o t have been d e t e c t e d a t the
A - l l intercomparison l e v e
f o r comparison, i s the "Suggeste
f o r Mn i n U.S. d r i n k i n g water ( 2 1 ) .
L o o k i n g beyond the q u e s t i o n o f d e t e c t i o n l i m i t s f o r a g i v e n
method-sample c o m b i n a t i o n , i t becomes i n t e r e s t i n g and i m p o r t a n t t o
c o n s i d e r d e t e c t i o n i n a l a r g e r sense, i e , d e t e c t i o n o f a phenomenon,
a s t a t e o f h e a l t h , an e n v i r o n m e n t a l p r o c e s s , e t c . , as m a n i f e s t by
chemical or r a d i o n u c l i d e composition. T h i s i m m e d i a t e l y r e q u i r e s us
to f a c e q u e s t i o n s o f s c i e n t i f i c u n d e r s t a n d i n g and modeling o f the
phenomenon under c o n s i d e r a t i o n . G i v e n t h a t , one can then address
the i s s u e o f the a p p r o p r i a t e s a m p l i n g d e s i g n and measurement t o
guarantee adequate d e t e c t i o n o f the phenomenon. I t i s a t t h i s l e v e l
t h a t d e t e c t i o n l i m i t concepts, a p p r o p r i a t e l y l i n k e d w i t h s c i e n t i f i c
u n d e r s t a n d i n g and i n s i g h t , can have t h e i r g r e a t e s t impact. To c i t e
one example, the b i o a s s a y o f Zn, a t r a c e element o f comparable
importance t o Fe f o r v e r t e b r a t e s ( 2 2 ) , can be q u i t e m i s l e a d i n g i n
the absence o f " b i o l o g i c a l i n s i g h t . " Iyengar ( 2 3 ) , f o r example, has
i l l u s t r a t e d how s a m p l i n g v a r i a t i o n s can generate q u i t e erroneous
c o n c l u s i o n s [ f a l s e p o s i t i v e and n e g a t i v e l e v e l s ] r e g a r d i n g t h i s
element depending on such b l o o d s a m p l i n g c o n d i t i o n s as f a s t i n g ,
i n t a k e o f normal f o o d , low Zn f o o d , h i g h Zn f o o d , s t r e s s , and
pregnancy. The l i n k between c h e m i c a l a n a l y s i s and the d e t e c t i o n o r
i n t e r p r e t a t i o n o f b i o l o g i c a l s t a t e s i s f u r t h e r confounded by complex
element i n t e r a c t i o n s . For example, due t o b i o a v a i l a b i l i t y v a r i a -
t i o n s , a g i v e n c o n c e n t r a t i o n o f "Cu i n a n i m a l f e e d s t u f f can be
d e f i c i e n t , adequate or t o x i c depending upon the Mo and S c o n t e n t o f
the f e e d s t u f f " ( 2 2 ) .
Our c o n c l u s i o n must be t h a t a sound c o n c e p t u a l b a s i s f o r
d e t e c t i o n l i m i t s i n c h e m i c a l and r a d i o l o g i c a l a n a l y s i s , l i n k e d w i t h
a sound s c i e n t i f i c u n d e r s t a n d i n g o f the r e s p e c t i v e measurement
p r o c e s s -- e s p e c i a l l y the n a t u r e o f the b l a n k , i s a p r e r e q u i s i t e t o
a d d r e s s i n g i m p o r t a n t s o c i e t a l q u e s t i o n s i n v o l v i n g b i o l o g i c a l and
e n v i r o n m e n t a l systems. I f we a r e t o g a i n f u l l b e n e f i t from such
measurements, however, we must work i n a m u l t i d i s c i p l i n a r y atmos-
phere, so t h a t we can d e s i g n our measurement p r o c e s s e s t o e x h i b i t
d e t e c t i o n (and q u a n t i f i c a t i o n ) l i m i t s w h i c h a r e adequate t o speak t o
these l a r g e r i s s u e s .

Acknowledgment
C o l l e a g u e s who s h a r e d i n f o r m a t i o n c o n c e r n i n g e x t a n t d e t e c t i o n l i m i t
p r a c t i c e s , and NRC and IAEA program needs, a r e g r a t e f u l l y acknowl
edged. I n c l u d e d i n t h i s p a r t i c u l a r acknowledgment a r e o t h e r
p a r t i c i p a n t s i n t h e IAEA C o n s u l t a n t s ' M e e t i n g ( s e e T a b l e I V ) , and
h o s t s f o r t h e NRC s i t e v i s i t s ( s e e p. 9, i n Ref. 1 2 ) . S p e c i a l
thanks go a l s o t o W. W. Meinke and V. Iyengar f o r many s t i m u l a t i n g
d i s c u s s i o n s and exchanges o f i n f o r m a t i o n , and t o F. R e i c h e l f o r
p e r m i s s i o n t o p r e s e n t some o f h e r f i n d i n g s i n F i g . 2.
Literature Cited
1. "Standard Radiological Effluent Technical Specifications for
Pressurized Water Reactors"; U. S. Nuclear Regulatory
Commission, NUREG-0472 Rev 3 September 1982
2. "Analytical Qualit
Runs, Certified Referenc
International Atomic Energy Agency, 1986-87.
3. Parr, R. M. "IAEA Biological Reference Materials"; In
"Biological Reference Materials: Availability, Uses, and Need
for Validation of Nutrient Measurement"; Wolf, W. R., Ed.;
Wiley: 1985; Chap. 3.
4. Keith, L. H.; Crummett, W.; Deegan Jr, J . ; Libby, R. .;
Taylor, J. K.; Wentler, G. "Principles of Environmental
Analysis," Analyt. Chem. 1983, 55, 2210.
5. Reichel, F. "Practical Estimation of the Limit of Detection in
Gamma Spectrometry Using a Commercially Available Program";
Chemistry Unit, IAEA Laboratories, 1986. See also K. Heydorn
in Ref. 6, pp. 179ff.
6. "Quality Assurance in Biomedical Neutron Activation Analysis";
IAEA, IAEA-TECD0C-323, 1984.
7. "Intercomparison of Cadmium and Other Elements in IAEA Horse
Kidney (H-8)"; IAEA, Progress Report No. 2, 1985.
8. Dybczynski R.; Veglia .; Suschny, O. "Report on the
Intercomparison Run for the Determination of Inorganic
Constituents of Milk Powder"; IAEA, IAEA/RL/68, 1980. See also
Ref. 3.
9. Iyengar, G. V.; Tanner, J. T.; Wolf, W. R.; Zeisler, R.
"Preparation of a Mixed Human Diet Material for the
Determination of Nutrient Elements, Selected Toxic Elements and
Organic Nutrients: a Preliminary Report"; submitted to The
Science of the Total Environment, 1986.
10. Parr, R. M. convenor, "IAEA Consultants, Meeting on Limit of
Detection," Vienna, December 1985.
11. Natrella, M. G. 'The Relation between Confidence Intervals and
Tests of Significance,' in Ku, H., Ed.; "NBS Spec Publ 300,"
1969.
12. Currie, L. A. "Lower Limit of Detection: Definition and
Elaboration of a Proposed Position for Radiological Effluent
and Environmental Measurements"; U. S. Nuclear Regulatory
Commission, NUREG/CR-4007, 1984.
13. Currie, L. .; DeVoe, J. R. 'Systematic Error in Chemical
Analysis,' in Devoe, J . R., Ed.; "Validation of the Measurement
Process"; American Chemical Society, Washington, 1977; Chap 3.

14. Currie, L. A. "Quality of Analytical Results, with Special

Reference to Trace Analysis and Sociochemical Problems"; Pure &
Appl. Chem. 1982, 54, 715. (See especially the discussion
surrounding Fig. 19.)
15. Iyengar, G. V.; Kasperek, K.; Feinendegen, L. E.; Wang, Y. X.;
Weese, H. "Determination of Co, Cu, Fe, Hg, Mn, Sb, Se and Zn
in Milk Samples"; The Science of the Total Environment. 1982,
24, 267.
16. Parr, R. M. "Technical Considerations for Sampling and Sample
Preparation of Biomedical Samples for Trace Element Analysis";
J. Res. NBS 1984 and reference 48 therein: Damsgaard; Heydorn,
K, RIS Rep Ris-M-1633, 1973.
17. "Quality Assurance in Biomedical Neutron Activation Analysis";
Analyt. Chim. Acta. 1984, 165, 1, IAEA, Report of an Advisory
Group.
18. Parr, R. M.; Houtermans H.; Schaerf K 'The IAEA
Intercomparison o
Spectra'; in "Computer
Spectroscopy"; U. S. Dept. of Energy, Sympos. Ser. 49, 1979,
544.
19. Currie, L. A. "The Limitations of Models and Measurements as
Revealed through Chemometric Intercomparison"; J. Res. NBS
1985, 90, 409.
20. Iyengar, G. V. "Concentrations of 15 Trace Elements in Some
Selected Adult Human Tissues and Body Fluids of Clinical
Interest from Several Countries"; Juelich Nuclear Research
Center, Rep. No. 1974, Feb 1985.
21. "Standard Methods for the Examination of Water and Wastewater";
13th Ed, Amer. Public Health Assoc., Washington, 1973, p. 33..
22. Iyengar, G. V. "Human Health and Trace Element Research"; Sci.
Total Environ. 1981, 19, 105.
23. Iyengar, G. V. "Normal Values for the Elemental Composition of
Human Tissues and Body Fluids: a New Look at an Old Problem,"
in "Trace Substances in Environmental Health - XIX"; Hemphill,
D. D., Ed.; Univ of Missouri, Columbia, 1985, 277.

Chapter 10
Effects of Analytical Calibration

Models on Detection Limit
Estimates
1
K. G. Owens , C. F. Bauer, and C. L. Grant
Department of Chemistry, University of New Hampshire,

Durham, NH 03824
Detection limi
bands around analytica highly
dependent on the experimental design and on the
statistical data treatment. Procedures are described
for testing the linearity of data and whether the
intercept differs s i g n i f i c a n t l y from zero.
Insensitivity of the correlation coefficient for the
evaluation of goodness of fit of calibration models
is emphasized, unweighted linear models with an
intercept often yield overly conservative detection
limits. Frequently, an unweighted zero-intercept
model is justified on both theoretical and
statistical grounds. This model yields confidence
bands and detection l i m i t s consistent with
experiment. When the variance of signal measurements
increases with concentration, more r e a l i s t i c
confidence bands and detection limits are produced by
weighting the data.
D e s p i t e numerous p a p e r s d e a l i n g w i t h the s p e c i f i c a t i o n of
a n a l y t i c a l method d e t e c t i o n l i m i t s (see, f o r example 1-6) much
r
disagreement remains about the choice of both e x p e r i m e n t a l and

computational procedures. In part, these disagreements appear to
be r e l a t e d to the t e c h n i c a l o b j e c t i v e s of the experimenter. For
example, a d e t e c t i o n l i m i t (DL) might be e s t i m a t e d from some
m u l t i p l e of the standard d e v i a t i o n of blank s o l u t i o n s i g n a l s
measured d u r i n g a short time i n t e r v a l u s i n g a p a i n s t a k i n g l y
optimized instrument. Such a DL estimate c l e a r l y provides useful
i n f o r m a t i o n but i t would be u n r e a l i s t i c to expect to m a i n t a i n an
e q u i v a l e n t DL d u r i n g an extended a n a l y s i s program w i t h r e a l
samples. Unfortunately, these differences are frequently ignored,
thereby encouraging unproductive controversy. It i s important to
remember that a DL i s not an i n t r i n s i c p r o p e r t y but r a t h e r , i t i s
Current address: Chemistry Department, Indiana University, Bloomington, IN 47405
0097-6156/88/0361 -0194$06.00/0

10. OWENS ET AL. Effects of Analytical Calibration Models 195
the i n t e r a c t i v e product of many variables including the method, the

instrumentation, the nature of the samples, and the experimenter.
f
T h i s d i s c u s s i o n w i l l be r e s t r i c t e d to e s t i m a t i o n of DL s
a p p r o p r i a t e f o r r o u t i n e a p p l i c a t i o n of methods i n which a
c a l i b r a t i o n function r e l a t e s a signal to concentration for a series
of standards. No attempt w i l l be made to evaluate the m e r i t s of
v a r i o u s terms such as l i m i t of q u a n t i t a t i o n , method d e t e c t i o n
l i m i t , lower l i m i t of r e l i a b l e assay measurement, and others. For
a d i s c u s s i o n o f those i s s u e s , the reader i s r e f e r r e d to the
a r t i c l e s c i t e d here and to other papers from t h i s symposium.
Consider a t y p i c a l procedure such as the spectrophotometric
determination of an analyte i n groundwater samples. Quite l i k e l y ,
a s i n g l e c a l i b r a t i o n curve w i l l be used to cover a c o n c e n t r a t i o n
range that extends from below the regulatory l i m i t (hopefully) to
some elevated concentration far removed from the l i m i t . In t h i s
s i t u a t i o n , a DL can be
c a l i b r a t i o n curve (7-14)
can then r e f l e c t the combined uncertainties i n sample analysis and
calibration.
C l e a r l y the design of the c a l i b r a t i o n procedure and the
s t a t i s t i c a l analysis of the data are both important considerations.
Questions which require attention are conveniently divided into two
groups; those p e r t a i n i n g to the e x p e r i m e n t a l design and those
pertaining to the s t a t i s t i c a l analysis. Design questions include:
a) What concentration range should be covered?

b) How many standards and blanks should be used?
c) How should standards be d i s t r i b u t e d over the range of
interest?
d) How many r e p l i c a t e measurements should be made on
standards and blanks, i n what o r d e r , and over what time
frame?
e) How should standards be prepared?
S t a t i s t i c a l analysis questions may include:

a) Are the signal measurements ( y ^ normally distributed at
each concentration l e v e l used?
2
b) Are the v a r i a n c e s o f the s i g n a l s ( S ) homogeneous,
y i
i.e., independent of concentration?

c) Is a l i n e a r model j u s t i f i e d or i s curvature indicated?
d) I f l i n e a r , i s the intercept (b ) s i g n i f i c a n t l y different
Q
from zero, i.e., i s a zero-intercept model suitable?

!
Calculation of CL s should proceed only after these questions
have been answered. Too f r e q u e n t l y , a l i n e a r model o f the
form y = b + Q i s f i t t e d using least squares procedures and
CLs are calculated without proper attention to these issues. If
the c a l i b r a t i o n model i s inappropriate, i t almost c e r t a i n l y follows
t h a t DL's based on C L ' s a r o u n d t h a t model a r e i n a c c u r a t e .
T h e r e f o r e , a major p o r t i o n of t h i s paper w i l l d e a l w i t h the
s e l e c t i o n of a c a l i b r a t i o n model. At the end of the paper, we
r e t u r n to a c o n s i d e r a t i o n o f the e f f e c t s of that model on DL
estimates.

Experimental Design Q u e s t i o n s .
C o n c e n t r a t i o n Range. Choice of the c o n c e n t r a t i o n range to be

covered i s g e n e r a l l y d i c t a t e d by the nature of the samples to be
analyzed, p r e c i s i o n and accuracy requirements, and the i n h e r e n t
l i m i t a t i o n s of the procedure. For example, an environmental
pollutant may be present i n samples at concentrations near the DL
on t h e f r i n g e s o f an a r e a but a t s u b s t a n t i a l l y e l e v a t e d
c o n c e n t r a t i o n s w i t h i n the most s e v e r e l y impacted area. Low
concentrations must be accurately estimated i n order to define the
geographic d i s t r i b u t i o n i n accordance with regulatory guidelines.
However, i t i s a l s o necessary to determine high c o n c e n t r a t i o n s
a c c u r a t e l y to plan cleanup s t r a t e g y . In these c i r c u m s t a n c e s , i t
may be best to produce two c a l i b r a t i o n curves; one f o r the low
concentration range from which the DL can be estimated, and one f o r
the f u l l c o n c e n t r a t i o
performed, i t i s l i k e l
Number of Standards, The number of standards required depends i n

l a r g e measure on the c o n c e n t r a t i o n range and the nature of the
anticipated functional relationship. A zero intercept l i n e a r model
for a l i m i t e d concentration range may be adequately defined with
three standards although most i n v e s t i g a t o r s p r e f e r f o u r or f i v e .
In contrast, a thorough evaluation of a c u r v i l i n e a r model requires
at least f i v e standards and more may be desiraole when spanning a
wide concentration range.
D i s t r i b u t i o n of Standards. The location of the c a l i b r a t i o n points

on the concentration axis exerts f a r more influence on DL estimates
than has been g e n e r a l l y recognized. Common p r a c t i c e i s to space
standards e q u i d i s t a n t a c r o s s t h e e n t i r e r a n g e o f i n t e r e s t .
However, for a specified number of standards, lower estimates of
f
DL s are obtained without compromising the r e l i a b i l i t y of the high
concentration range when an unsymmetrical d i s t r i b u t i o n favoring low
c o n c e n t r a t i o n s i s used. These e f f e c t s w i l l be i l l u s t r a t e d i n a
l a t e r section where CL computations are discussed.
R e p l i c a t i o n of S t a n d a r d s . An obvious b e n e f i t of r e p l i c a t i o n i s
improved r e l i a b i l i y of the results. Other benefits are the ease of
t e s t i n g the goodness of f i t of the c a l i b r a t i o n model and the
o p p o r t u n i t y to i n t e r s p e r s e measurements on standards randomly
a c r o s s an e n t i r e l o t of samples to which that c a l i b r a t i o n curve
w i l l apply. By t h i s arrangement the standard deviation estimated
from the c a l i b r a t i o n data w i l l usually correspond c l o s e l y to the
value estimated from r e p l i c a t e measurements on samples. Otherwise,
r e p r o d u c i b i t y of samples may be poorer than f o r standards. Of
course, t h i s assumes no systematic d r i f t of signal response during
the course of the measurements. I f d r i f t i s suspected, i t can be
checked by making s e v e r a l measurements on standards over an
extended time period. Any procedure should be demonstrated to be
p e r f o r m i n g n o r m a l l y and i n c o n t r o l before s t a r t i n g on a l o t of
samples.
The question of how many r e p l i c a t e measurements to make must
i n c l u d e c o n s i d e r a t i o n of the magnitude of v a r i a b i l i t y , the time

r e q u i r e d , the c o s t of each measurement, and the r e l i a b i l i t y

r e q u i r e d i n the f i n a l r e s u l t . For many s i t u a t i o n s we find
duplicates or t r i p l i c a t e s are quite adequate.
Preparation of Standards, To meet the requirement of independence

of e r r o r s , each r e p l i c a t e and each standard should be prepared
s e p a r a t e l y and w i t h great care (15). Subsequent r e g r e s s i o n
a n a l y s i s w i l l assume no e r r o r i n the c o n c e n t r a t i o n (or at l e a s t ,
that the e r r o r i s s m a l l i n comparison to the e r r o r i n s i g n a l
measurement) and that the errors are independent. Thus, solution
standards should i d e a l l y be prepared from more than one stock
solution using a v a r i e t y of pipets and volumetric flasks.
Choosing a C a l i b r a t i o n Model.
An assumed r e g r e s s i o
Choosing the c o r r e c t
results. Although several recent papers (16-18) e x t o l l the v i r t u e s
of nonlinear c a l i b r a t i o n curves as a means of improving accuracy or
to extend the range of concentrations covered, t h i s discussion w i l l
consider only l i n e a r models. Evaluation of nonlinear models i s an
extension of the l i n e a r case with s i m i l a r conceptual framework.
unweighted l e a s t squares curve f i t t i n g i s based on the
a s s u m p t i o n s t h a t (a) measurement e r r o r s f o l l o w a Gaussian
d i s t r i b u t i o n and t h a t (b) v a r i a n c e s a r e i n d e p e n d e n t o f
concentration, i . e . , t h e y a r e homogeneous. In t y p i c a l
calibrations, i n s u f f i c i e n t data are c o l l e c t e d to r i g o r o u s l y t e s t
e i t h e r assumption. F o r t u n a t e l y , modest v i o l a t i o n s do not cause
serious errors but Garden et a l . (19) warned that incorrect use of
an unweighted l e a s t squares a n a l y s i s could cause gross e r r o r s i n
the estimation of trace concentrations. In the i n i t i a l portion of
t h i s discussion we w i l l consider examples where both assumptions
appear v a l i d . L a t e r we w i l l examine the e f f e c t s of nonuniform
variance.
Linear Model W i t h I n t e r c e p t . There are two d i s t i n c t l i n e a r f i r s t -

order r e g r e s s i o n models t h a t a r e g e n e r a l l y e n c o u n t e r e d i n
analytical calibration. The non-zero intercept model i s the most
f a m i l i a r , and i t i s given by Equation 1.
y = b Q + b x (1)
The estimates of intercept (b ) and slope (b^) are calculated so as

Q
to m i n i m i z e the sum of squares (SS) of the d e v i a t i o n s of the

o b s e r v e d s i g n a l s (y^) f r o m t h e p r e d i c t e d v a l u e (y) a t any
c o n c e n t r a t i o n (x) without c o n s t r a i n t s . For some d e t e r m i n a t i o n s ,
however, theory predicts that the response of the instrument should
be l i n e a r with concentration and should also be zero when there i s
no a n a l y t e present. Thus, i f the instrument has been c a l i b r a t e d
c o r r e c t l y , the c a l c u l a t e d l i n e should pass through the o r i g i n by
d e f i n i t i o n . The proper r e g r e s s i o n model would then be the zero
intercept model shown as Equation 2.
y = b,
x
(2)
o

The e s t i m a t e o f the true s l o p e , b 1 o , i s c a l c u l a t e d so as to

minimize the SS of deviations from the l i n e with the r e s t r i c t i o n
that the l i n e must pass through the o r i g i n . Each of these models
w i l l be considered i n the following paragraphs.
To f a c i l i t a t e t h i s discussion, we have fabricated some t y p i c a l
spectrophotometric c a l i b r a t i o n data employing duplicate absorbance
measurements at five concentrations. A reagent blank was used to
set zero absorbance. Two s e t s o f data are shown i n Table I. The
absorbance v a l u e s at each o f the 4 lowest c o n c e n t r a t i o n s are
i d e n t i c a l f o r each set. The d i f f e r e n c e occurs i n the absorbance
values f o r the h i g h e s t standard where a negative d e v i a t i o n from
Beers Law i s represented by a reduced absorbance for case II. The
regression equations and c o r r e l a t i o n c o e f f i c i e n t s were calculated
according to standard equations available i n any text on regression
analysis.
Table I. Dat
for Spectrophotometri
Concentrations Measured Absorbances (v)

o f Standards (x) Caag_i Case I I
0.500 0.054, 0.050 0.054, 0.050
1.00 0.103, 0.109 0.103, 0.109
2.00 0.202, 0.192 0.202, 0.192
5.00 0.494, 0.514 0.494, 0.514
10.00 0.975, 1.005 0.915, 0.945
Least Squares Model

With Intercept y=0.00431+0.0988x y=0.0149+0.0927x
Correlation Coefficients r = 0.9996 r = 0.9988
Models Through Origin y = 0.0994x y = 0.0948x
Goodness of F i t . The f i t t e d model w i t h i n t e r c e p t f o r Case I i s

seen to have a c o r r e l a t i o n coefficient of 0.9996 which would often
be i n t e r p r e t e d to mean that the equation f i t s the data very w e l l .
However, we s h a l l see f r o m the Case I I d a t a s e t t h a t t h e
c o r r e l a t i o n c o e f f i c i e n t i s not a s e n s i t i v e method of e v a l u a t i n g
curve f i t . H u n t e r (2) n o t e s t h a t i n s t a t i s t i c a l t h e o r y ,
c o r r e l a t i o n i s a measure of the r e l a t i o n s h i p between two random
(dependent) v a r i a b l e s . In a c a l i b r a t i o n problem, however, i t i s
assumed that there i s a d e f i n i t e functional relationship between
the dependent and independent v a r i a b l e s . Correlation, in i t s
s t r i c t s t a t i s t i c a l sense, does not exist. Van Arendonk et a l . (21)
point out that the c o r r e l a t i o n coefficient i s an i n s e n s i t i v e t o o l
for use i n e v a l u a t i n g the q u a l i t y of the f i t t e d e q u a t i o n , and i t s
use i n such a manner may lead to erroneous conclusions.
We b e l i e v e that i t i s f a r more i n s t r u c t i v e to perform a
regression a n a l y s i s i n which the s o u r c e s o f v a r i a t i o n are
f r a c t i o n a t e d i n t o the sums of squares (SS) a t t r i b u t a b l e to
regression and the SS for residuals. When r e p l i c a t e measurements
have been made, the residual SS can be further fractionated i n t o a

s y s t e m a t i c e r r o r component and a random e r r o r component. The

systematic error component i s designated the SS due to lack of f i t
(LOF) b e c a u s e i t a r i s e s from t h e inadequacy of the f i t t e d
regression model to describe the experimental points. Table I l - a
g i v e s the e q u a t i o n f o r c a l c u l a t i n g the SS of r e s i d u a l s w i t h N-2
degrees of freedom (d.f.), since two r e g r e s s i o n c o e f f i c i e n t s were
fitted. Many s t a t i s t i c a l analysis programs routinely provide the
SS of residuals. The SS for random error (SS error) i s independent
of the r e g r e s s i o n model employed, i . e . , i t depends s o l e l y on the
d i s t r i b u t i o n o f r e p l i c a t e s around the mean response f o r each
standard. When duplicate measurements have been acquired for each
standard, the SS error i s calculated as shown i n Table - a where
d s d i f f e r e n c e i n s i g n a l f o r each set o f d u p l i c a t e s . T h e t o t a l d.f.
i n t h i s e r r o r e s t i m a t e would be equal to the number of d u p l i c a t e
sets since each would contribute 1 d.f. The SS for LOF i s obtained
by d i f f e r e n c e between
S i m i l a r l y , the d.f. associate
These c a l c u l a t i o n s are i l l u s t r a t e d i n Table I I f o r l i n e a r
models w i t h i n t e r c e p t s f i t t e d to the data s e t s of Table I.
I n s p e c t i o n of Table II r e v e a l s t h a t the F - r a t i o f o r LOF f o r the
Case I r e s u l t s i s not s i g n i f i c a n t as expected, i . e . , the model i s
an adequate d e s c r i p t i o n of the data. For the Case I I r e s u l t s ,
however, the LOF i s s i g n i f i c a n t at the 0.10 s i g n i f i c a n c e l e v e l
despite finding a c o r r e l a t i o n coefficient of 0.9988! With the high
p r o b a b i l i t y that the l i n e a r model does not p r o p e r l y f i t the data
for Case I I , i t seems unreasonable to use such a c a l i b r a t i o n curve
without t r y i n g to r e s o l v e the problem. Note that the nature of
t h i s test i s such that the LOF w i l l not show significance i f large
random errors are present. In fact, when random error i s large, i t
i s d i f f i c u l t to detect systematic variations that might r e s u l t i n
LOF. In t h i s example, however, random e r r o r i s the same f o r each
case. The LOF i s caused by a negative deviation of absorbance for
the highest concentration standard. The problem can be resolved by
r e d u c i n g the c o n c e n t r a t i o n of the h i g h e s t standard to the upper
l i m i t of the l i n e a r range or possibly by f i t t i n g a nonlinear model.
The important point i s that the c o r r e l a t i o n coefficient provides no
r e a l insight concerning the extent or nature of residuals whereas
the LOF test does.
I t i s i m p o r t a n t to note t h a t an o b s e r v a t i o n (or set o f
o b s e r v a t i o n s ) on a standard may be r e j e c t e d as an o u t l i e r o n l y i f
i t i s not at the extreme ends of the c a l i b r a t i o n curve. I f the
lowest (or highest) standard appears to be an o u t l i e r , i t can not
be determined from the data collected whether the concentration of
the standard i s i n e r r o r or i f the response o f the i n s t r u m e n t i s
beginning to d e v i a t e from l i n e a r i t y . The " o u t l y i n g " o b s e r v a t i o n
would have to be retained unless additional measurements made on a
standard o f lower (higher) c o n c e n t r a t i o n i n d i c a t e s t h a t the
deviation from the calculated l i n e i s not due to n o n - l i n e a r i t y of
the response function.
When r e p l i c a t e measurements are not a v a i l a b l e , a thorough
analysis i s required of the residuals: the i n d i v i d u a l differences
between the e x p e r i m e n t a l p o i n t s and the c a l c u l a t e d r e g r e s s i o n
l i n e . P a t t e r n s i n r e s i d u a l p l o t s p r o v i d e i n s i g h t c o n c e r n i n g the
v a l i d i t y of the f i t t e d equation and possible causes when the f i t i s

Table -a. Formulation of Regression Analysis Table

Using The Calibration Data of Table I and
A Linear Model With Intercept
Degrees of Mean
Source of Sum of squares freedom square F-ratio
Variation (SS) (df ) (MS) (F)
Residual |y - 1 ^ 3
2
.
Error 5 SS error
2 5
Lack of F i t Residual SS - Error SS 3 SS LOF MS L0F

(LOF) 3 MS error
Table -b. Regression Analyses With LOF Tests For

Table I Calibration Data Using Intercept Model
Source of
Variation SS df MS F-ratio*
Case I
Residual 0.000872 8 0.000109
Error 0.000726 5 0.000145
LOF 0.000146 3 0.0000487 0.34
Case I I
Residual 0.002518 8 0.000315
Error 0.000726 5 0.000145
LOF 0.001792 3 0.000597 4.12
*The F - r a t i o s r e q u i r e d f o r 3 and 5 df at v a r i o u s s i g n i f i c a n c e
l e v e l s are 3.62 f o r 0.10, 5.41 f o r 0.05.

poor. T h i s s u b j e c t i s comprehensively d i s c u s s e d i n Draper and

Smith (15.).
Zero I n t e r c e p t Model. For s p e c t r o p h o t o m e t r i c d e t e r m i n a t i o n s ,

theory p r e d i c t s that response of the i n s t r u m e n t should be l i n e a r
with concentration and that the response should be zero when there
i s no a n a l y t e present. The zero i n t e r c e p t r e g r e s s i o n model
( E q u a t i o n 2) p r o v i d e s p a r a m e t e r e s t i m a t e s which meet t h i s
restriction. The e x p r e s s i o n used to c a l c u l a t e the slope of the
l i n e through the o r i g i n i s :
*1 = ^ (3)

F i t t e d models through the o r i g i n are shown i n Table I for the two
sets of data previously discussed.
Before the equatio
intercept model i s employe
to determine i f the model i s adequate to describe the experimental
data. Regression analysis tables are constructed p r i o r to testing
the s t a t i s t i c a l v a l i d i t y of the assumption that the i n t e r c e p t of
the l i n e i s zero. The format f o r c a l c u l a t i o n of the r e g r e s s i o n
a n a l y s i s t a b l e s i s shown i n Table I l l - a and the analyses of the
Table I data are shown i n Table I l l - b .
Inspection of these tables shows that the LOF test r e s u l t s are
very s i m i l a r to those for the models with intercepts. Comparison
of Tables - b and - b reveals that the SS residuals are somewhat
l a r g e r f o r the zero i n t e r c e p t models than f o r the models w i t h an
intercept. This difference can be used to test the hypothesis that
the intercept i s zero. F i r s t , i t must be demonstrated that the LOF
i s not s i g n i f i c a n t s i n c e i t would not make good sense to t e s t the
zero i n t e r c e p t h y p o t h e s i s f o r l i n e a r models shown not to f i t the
data. Furthermore, the SS error and SS(LOF) should not be combined
as SS residuals when LOF i s s i g n i f i c a n t . These requirements are
met by the Case I r e s u l t s . To t e s t the h y p o t h e s i s that the
intercept does not d i f f e r s i g n i f i c a n t l y from zero, calculate:
(SS residual for zero (SS residual of model

intercept model) - with intercept)
F = (4)
MS residual of model with intercept
For the Case I d a t a , F = Q,QQv96v-Q,QQQ872 .81

= 0
0.000109
The d.f. i n the numerator w i l l always be 1 because (N-1-(N-2)=1
and, t h e r e f o r e the d i f f e r e n c e i n these SS are d i v i d e d by 1 to get
the MS. The d.f. i n the denominator are N-2 or 8 i n t h i s example.
At the 0.05 s i g n i f i c a n c e l e v e l , the r e q u i r e d F value w i t h 1 and 8
d.f. i s 5.32. C l e a r l y , we can not r e j e c t the h y p o t h e s i s that the
intercept i s zero and consequently we conclude that t h i s model i s
consistent with the data.
It can be very advantageous to achieve a c a l i b r a t i o n that has
a zero intercept i f i t can be demonstrated that t h i s condition can
be sustained on a long term basis. We find that some systems that
are c a r e f u l l y zeroed on blanks w i l l meet t h i s requirement. Under

Table I l l - a . Formulation of Regression Analysis Table

Using The C a l i b r a t i o n Data i n Table I and a
Zero Intercept Model
Degrees of Mean
Sum of squares freedom square F-ratio
(SS) / <df) (MS) (F)
Residual
Z2
Error d
5 SS error
5
Lack of F i t Residual SS-Error SS 4 as LQF MS LQF
4 MS error
Table I l l - b . Regression Analyses With LOF Tests

For Table I C a l i b r a t i o n Data Using Zero Intercept Model
Source of
Variation _S_ JIS. F-ratlo
Case I
Residual 0.000960 0.000107
Error 0.000726 0.000145
LOF 0.000234 0.0000585 0.40
Case I I
Residual 0.003577 0.000397
Error 0.000726 0.000145
LOF 0.002851 0.000713 4.92
*The F - r a t i o s r e q u i r e d f o r 4 and 5 d.f a t v a r i o u s s i g n i f i c a n c e

l e v e l s are 3.52 f o r 0.10, 5.19 f o r 0.05.

such conditions, c a l i b r a t i o n for each l o t of samples to be analyzed

requires only that r e p l i c a t e s of the highest standard be run. The
mean of duplicate or t r i p l i c a t e measurements of instrument response
i s checked t o see i f i t f a l l s w i t h i n the c o n f i d e n c e i n t e r v a l s
established f o r the o r i g i n a l c a l i b r a t i o n curve. When the mean i s
w i t h i n the i n t e r v a l s (as i t w i l l be most o f the time), the system
i s considered to be i n control and the o r i g i n a l c a l i b r a t i o n curve
i s employed. I f the mean i s o u t s i d e the i n t e r v a l s , f u r t h e r
instrument c a l i b r a t i o n (adjustments) must be made t o r e t u r n the
response t o a s t a t e o f c o n t r o l before a n a l y s i s o f samples i s
attempted.
Although theory predicts that the calculated l i n e should pass
through the o r i g i n by d e f i n i t i o n , sometimes the experimental data
i n d i c a t e that the zero i n t e r c e p t model i s not adequate, i.e., the
intercept of the l i n e i s s t a t i s t i c a l l y d i f f e r e n t from zero. When
t h i s happens, valuable
method i s available. Fo
one described here, a positiv intercep y presenc
uncorrected background interferences or that the data have not been
adjusted for the blank, undetected nonlinearity can also cause the
i n t e r c e p t t o d e v i a t e from zero. Another p o s s i b i l i t y i s i f the
c o n c e n t r a t i o n s o f c a l i b r a t i o n standards a r e high by a constant
amount, i t w i l l s h i f t the c a l i b r a t i o n curve to the l e f t and produce
a positive intercept. I f the intercept of the calculated equation
i s negative, i t may indicate that the concentrations are a c t u a l l y
lower than what i s calculated. Thus knowledge o f the method being
used, coupled w i t h a thorough s t a t i s t i c a l a n a l y s i s o f the data,
could indicate chemical problems with the method that caused i t to
deviate from theory.
Heterogeneity o f V a r i a n c e s . E a r l i e r i n t h i s d i s c u s s i o n , we
promised t o r e t u r n t o the q u e s t i o n o f non-uniform v a r i a n c e o f
signal measurements over the concentration range used. Often i t i s
not clear by inspection whether variances are heterogeneous. One
way t o t e s t t h i s assumption i s by B a r t l e t t ' s Chi-square t e s t .
Because the calculations f o r t h i s test are quite extensive we use a
s i m p l e r t e s t based on the comparison o f ranges (22) The t e s t
i n v o l v e s c a l c u l a t i n g the range between the h i g h e s t and lowest
responses r e p o r t e d f o r each standard and c a l c u l a t i n g the r a t i o
R / ( R | + R + R ) where R-j, R R are the ranges f o r each o f k
max 2 k 2 k
standards and R m a x i s the largest range. This test can be applied

with as few as two r e p l i c a t e s a t each concentration, but i t i s more
r e l i a b l e and more sensitive when a greater amount o f r e p l i c a t i o n i s
available. When the calculated r a t i o exceeds the tabular value f o r
a 0.05 s i g n i f i c a n c e l e v e l , we r e j e c t the h y p o t h e s i s t h a t the
variances are homogeneous.
A variety of approaches have been recommended to deal with the
problem o f h e t e r o g e n e i t y o f v a r i a n c e s . We f a v o r the use o f
weighting because "the accuracy o f the c a l i b r a t i o n curve i s almost
i n v a r i a b l y i n c r e a s e d when w e i g h t i n g f a c t o r s a r e i n c o r p o r a t e d ,
taking i n t o account the experimentally determined variances at each
measurement p o i n t " (17). Weighting can be approached u s i n g
d i f f e r e n t degrees of sophistication, but a l l methods are based on
obtaining estimates f o r the variance across the measurment range.

204 DETECTION INANALYTICAL CHEMISTRY
For i n s t a n c e , Garden et a l . (_U1) recommend that a f u n c t i o n a l

relationship be f i t t e d to plots of variance and standard deviation
versus the independent v a r i a b l e ( c o n c e n t r a t i o n ) . A weighting
procedure we often f i n d u s e f u l when a l a r g e number of r e p l i c a t e s
are available i s an empirical one i n which the weighting factor i s
s i m p l y the r e c i p r o c a l o f the v a r i a n c e ( H ) . T h i s scheme g i v e s
lower emphasis to more highly variable observations. The change i n
slope of the f i t t e d model i s usually quite minimal compared to an
unweighted e s t i m a t e , but there i s o f t e n a l a r g e r e d u c t i o n i n the
standard deviation and consequently i n the DL.
Equations used for the c a l c u l a t i o n of regression c o e f f i c i e n t s
and f o r s t a t i s t i c a l a n a l y s e s are the same as f o r unweighted data
except that each datum i s modified through m u l t i p l i c a t i o n by the
appropriate weighting factor. (It could be argued that a weighting
f a c t o r of 1.0 i s used throughout when p e r f o r m i n g unweighted
calculations). These
variety of software programs
Draper and Smith (15) an Oppenheime (II)
Another recommended approach to deal with t h i s problem i s the
use of transformations, especially the log transform ( J ) . In our
hands, t h i s procedure has been less satisfactory than weighting but
i t represents an alternative approach deserving of consideration.
Confidence L i m i t s and Detection L i m i t s ,
The widths of confidence i n t e r v a l s around c a l i b r a t i o n curves depend

not only on the v a r i a b i l i t y i n the data, but also on the regression
model chosen ( U ) . For the non-zero i n t e r c e p t model, the best
a b s o l u t e p r e c i s i o n occurs at x, y which i s the c e n t r o i d of the
r e g r e s s i o n l i n e ( F i g u r e 1a). T h i s c i r c u m s t a n c e p e r t a i n s l a r g e l y
because x, y i s the a x i s of r o t a t i o n f o r the u n c e r t a i n t y i n the
f i t t e d slope of the c a l i b r a t i o n curve. In c o n t r a s t , the best
a b s o l u t e p r e c i s i o n f o r the zero i n t e r c e p t model o c c u r s when the
concentration (x) i s zero (Figure 1b).
One of the commonly used methods for estimation of a DL with a
non-zero intercept model i s that of Hubaux and Vos ( ) Equations
for c a l c u l a t i n g the CL's and the DL are g i v e n i n the o r i g i n a l
paper. In Figure 1a, the point where the upper band i n t e r s e c t s the
Y a x i s i s designated Y . D L Below t h i s s i g n a l , an i n d i v i d u a l
measurement cannot be d i s t i n g u i s h e d from zero based on the
c a l i b r a t i o n data c o l l e c t e d . According to Hubaux and Vos, the
lowest concentration that can be distinguished from zero i s given
by the intersection of the horizontal constructed at Y D L with the
lower confidence band, designated X i n Figure 1a.
n L
S e v e r a l problems a r r i s e when t h i s procedure i s f o l l o w e d .

Because the c e n t r o i d of the c a l i b r a t i o n data i s f a r removed from
f
z e r o , t h e C L s d i v e r g e g r e a t l y n e a r z e r o due t o t h e l o n g
e x t r a p o l a t i o n a s s o c i a t e d w i t h u n c e r t a i n t y i n the slope.
Furthermore, i f the variance tends to increase with concentration,
the pooled standard deviation estimate i s i n f l a t e d r e l a t i v e to the
a c t u a l standard d e v i a t i o n near the DL. As a consequence, the
e s t i m a t e d DL i s f r e q u e n t l y above the lowest standard. B a i l e y et
&
1 ( 2 3 ) reported recoveries up to 97% for spiked concentrations of

Figure 1. Confidence L i m i t s For (a) Unweighted Non-Zero

Intecept, (b) Unweighted Zero I n t e r c e p t , and (c)
Weighted Non-Zero I n t e r c e p t Regression Models.
Y i s t n e m i n i m u m
DL d e t e c t a b l e s i g n a l and X is
D L
the concentration detection l i m i t .

dye that were below the DL calculated by t h i s procedure! Clearly

t h i s method of estimating the DL i s overly conservative.
Hubaux and Vos r e c o g n i z e d t h a t when the v a r i a n c e i n c r e a s e d
w i t h c o n c e n t r a t i o n , an i n f l a t e d e s t i m a t e of the DL would r e s u l t .
Their suggestion to cope with t h i s problem was to recommend that
most of the r e p l i c a t i o n be conducted w i t h the low c o n c e n t r a t i o n
standards, a procedure that i s somewhat analogous to weighting. In
extreme cases, the high standards are not replicated at a l l . Such
an approach w i l l n o r m a l l y reduce the s i z e of the standard
d e v i a t i o n and i t a l s o moves the x, y p o i n t much c l o s e r to zero,
thereby r e d u c i n g the l e n g t h of e x t r a p o l a t i o n . P r e d i c t a b l y , DL
estimates drop when t h i s i s done but the quality of c a l i b r a t i o n at
higher concentrations also suffers.
In our experience, the unweighted z e r o - i n t e r c e p t model i s
often j u s t i f i e d . As shown i n F i g u r e 1 b , CLs f o r t h i s model
r e f l e c t our i n t u i t i v e expectation
c o n s i s t e n t w i t h our e x p e r i m e n t a
c a r e f u l study of the s u i t a b i l i t y of t h i s model as d e s c r i b e d
earlier.
When a non-zero intercept model i s required and when variances
are not homogeneous, then weighted procedures are favored.
f
Equations for the c a l c u l a t i o n of CL s using weighting are given by
Oppenheimer et a l . ( 1 1 ) . The general shape of these bands i s shown
i n Figure 1 c . Because the weighted centroid X , Y i s much closer
w w
to the o r i g i n than f o r unweighted data, and because the weighted

s t a n d a r d d e v i a t i o n e s t i m a t e r e f l e c t s more s t r o n g l y t h e
r e p o d u c i b i l i t y at low concentrations, the DL l i m i t estimates are
substantially lower than f o r unweighted data. In our experience,
estimates derived i n t h i s manner more accurately r e f l e c t the r e a l
c a p a b i l i t y of an a n a l y t i c a l method.
Summary,
Careful experimental design i s shown to be an essential prelude to

analytical calibration. Such experiments lend themselves to
thorough s t a t i s t i c a l evaluation and to improved confidence i n the
experimental estimates derived therefrom. Large variations i n D.L.
estimates can be explained by differences i n c a l i b r a t i o n models and
the associated assumptions and computations.
Acknowledge^
Thanks are due to Mr. Thomas F. J e n k i n s , US Army Cold Regions

Research and E n g i n e e r i n g Laboratory, Hanover, NH f o r numerous
helpful discussions of the concepts i n t h i s paper.
Literature Cited.
1. Currie, L.A. Anal. Chem. 1968, 40, 586-93.

2. Kaiser,H. Two Papers on the Limit of Detection of a
Complete Analytical Procedure; Hafner: New York, 1969.
3. Boumans, P.W.J.M. Spectrochim. Acta 1978, 33B, 625-34.
4. Glaser, J..; Foerst, D.L.; McKee, G.D.; Quave, S.A.; Budde,
W.L., Environ. Sci. Technol. 1981, 15, 1426-35.

5. Winefordner, J . D . ; Long, G.L. A n a l . Chem. 1983, 55, 712A-24A.

6. Kirchmer, C . J . Environ. Sci Technol. 1983, 17, 174A-81A.
7. L i n n i g , F.J.; Mandel, J. Anal. Chem. 1964, 36, 25A-32A.
8. Hubaux, . ; Vos, G. Anal. Chem. 1970, 42, 849-55.
9. Schwartz, L . M . Anal. Chem. 1977, 49, 2062-68.
10. Agterdenbos, J. Anal. Chim. Acta. 1979, 108, 315-23.
11. Agterdenbos, J ; Maessen, F.J.M.J.; Balke, J . Anal. Chim. Acta.
1981, 132, 127-37.
12. Kurtz, D.A. Anal. Chim. Acta. 1983, 150, 105-14.
13. Oppenheimer, L.; C a p i z z i , T.P.; Weppelman, R . M . ; Mehta, H.
Anal. Chem., 1983, 55, 638-43.
14. Schwartz, L . M . Anal. Chem., 1986. 58, 246-50.
15. Draper, N.R.; Smith, H. Applied Regression Analysis, 2nd Ed.;
John Wiley & Sons, I n c . , New York, 1981
16. Schwartz, L . M . Anal. Chem. 1983, 55, 1424-26.
17. Jonckheere, J..; De Leenheer, A . P . ; S t e y a e r t , H.L. Anal.
Chem. 1983, 55
18. Barnett, W.B. Spectrochim. Acta 1984, 39B, 829-39.
19. Garden, J.S.; M i t c h e l l , D . S . ; M i l l s , W.M. Anal. Chem.
1980, 52, 2310-15.
20. Hunter, J.S. J. Assoc. Off. Anal. Chem. 1981, 64 574-83.
21. Van Arendonk, M . D . ; Skogerboe, R . K . ; Grant, C . L . Anal. Chem.
1981, 53, 2349-50.
22. Pearson, E.S.; Hartley, H . O . , Eds. Biometrika Tables for
S t a t i s t i c i a n s , Cambridge Univ. Press, 1966, Table 31b.
23. B a i l e y , C . J . ; Cox, E . A . ; S p r i n g e r , J.A. J. Assoc. Off. A n a l .
Chem. 1978, 61, 1404-14.

Chapter 11
Critical Assessment
of Detection Limits
for Ion Chromatography
William F. Koch and Walter S. Liggett
National Bureau of Standards, Gaithersburg, MD 20899
The s t a t i s t i c a l basis for ion chromatography detection

limits is investigate
chromatograms by
methods reveal two important chromatogram noise
components, a cyclic variation caused by the pump and
some large low-frequency variations with obscure
origins. The component due to the pump can be removed
from the chromatogram. The causes of the low frequency
component should be investigated because these causes
may not satisfy the prerequisites of s t a t i s t i c a l
inference. Detection limit assessment depends on the
choice of a peak detection algorithm. This algorithm
must include a method for separating the low frequency
component from the peak of interest and the a method
for locating the peak in time. Algorithms that search
for the peak in time cannot be assessed in the same way
as algorithms that involve no search. This difference
is discussed.
The d e t e c t i o n l i m i t i n d i c a t e s the performance o f an i n s t r u m e n t a t low

a n a l y t e c o n c e n t r a t i o n s . T h i s i n d i c a t i o n may be used as a g u i d e t o
i n s t r u m e n t o p t i m i z a t i o n , as a gauge o f t h e s u i t a b i l i t y o f an
i n s t r u m e n t f o r a p a r t i c u l a r a p p l i c a t i o n , o r as a c r i t e r i o n f o r t h e
i n t e r p r e t a t i o n o f low c o n c e n t r a t i o n measurements. T h i s paper
c o n c e n t r a t e s on t h e l a t t e r u s e o f d e t e c t i o n l i m i t s and expands t h e
d i s c u s s i o n t o i n c l u d e a l l a s p e c t s o f s t a t i s t i c a l i n f e r e n c e on low
c o n c e n t r a t i o n measurements. I n t h i s c a s e , the use o f t h e d e t e c t i o n
l i m i t i s c o n f i n e d t o the measurements i n q u e s t i o n and t o t h e s t u d y a t
hand. The u s e o f d e t e c t i o n l i m i t s f o r i n s t r u m e n t o p t i m i z a t i o n and
f o r s u i t a b i l i t y judgments r e q u i r e s a broader p e r s p e c t i v e t h a t c o v e r s
the v a r i o u s c o n d i t i o n s under which the i n s t r u m e n t might be used.
The q u e s t i o n o f what d e t e c t i o n l i m i t i s a c h i e v e d i n the course
of a s e t o f measurements i s b o t h c o m f o r t i n g l y s p e c i f i c and v e r y
demanding. The q u e s t i o n i s s p e c i f i c i n t h a t o n l y the o p e r a t i n g
c o n d i t i o n s used f o r t h e s e t o f measurements need t o be c o n s i d e r e d .
F u r t h e r , good p r a c t i c e s u g g e s t s t h a t t h e s e o p e r a t i n g c o n d i t i o n s be
T h i s chapter not subject to U.S. copyright

P u b l i s h e d 1988 A m e r i c a n C h e m i c a l Society

11. KOCH & LIGGETT Critical Assessment of Detection Limits 211
l i m i t e d as much as p o s s i b l e . Thus, the amount o f d a t a t h a t i s needed

i s r e l a t i v e l y modest. The a p p r o p r i a t e d a t a c o u l d be c o l l e c t e d as
p a r t of the l a b o r a t o r y q u a l i t y a s s u r a n c e program. The q u e s t i o n i s
demanding i n t h a t c o n c l u s i o n s w i t h b e l i e v a b l e p r o b a b i l i t i e s of e r r o r
are o f t e n needed. T h i s means t h a t the measurements used t o d e r i v e
p r o p e r t i e s of the e r r o r must have the same e r r o r p r o p e r t i e s as the
measurements about w h i c h c o n c l u s i o n s are t o be drawn. I n o t h e r
words, the unknown samples and the q u a l i t y c o n t r o l samples must b o t h
r e s u l t from a measurement p r o c e s s t h a t i s under c o n t r o l . I n many
c a s e s , t h i s c o n d i t i o n i s not e a s i l y a c h i e v e d because of s u b t l e
d i f f e r e n c e s between the c o n d i t i o n s under w h i c h the q u a l i t y c o n t r o l
measurements are made and the c o n d i t i o n s under w h i c h the r e a l
measurements are made. S p e c i f i c causes of such d i f f e r e n c e s , sample-
to-sample c a r r y o v e r and m e c h a n i c a l t r a n s i e n t s , are d i s c u s s e d below.
T h i s paper c o n s i d e r s i o n chromatography, w h i c h i s a form o f
l i q u i d chromatography base io separatio
f o l l o w e d by c o n d u c t i m e t r i
p a r t i c u l a r , t h i s paper i
(See t h e d i s c l a i m e r . ) T h i s i n s t r u m e n t was s e t up f o r the measurement
o f n i t r a t e and s u l f a t e a t c o n c e n t r a t i o n s below 1 mg/L. Except i n the
cases noted below, the i n s t r u m e n t was c o n f i g u r e d as f o l l o w s : The
a n i o n s e p a r a t o r column was number AS4A ( D i o n e x ) ; the e l u e n t was an
a d m i x t u r e of 0.75 mmol/L NaHCC^ and 2.0 mmol/L Na2C3; the f l o w r a t e
was 2.0 mL/min; and the sample loop volume was 20 pL. The background
conductance o f the e l u e n t was c h e m i c a l l y suppressed w i t h a h o l l o w
f i b e r c h e m i c a l s u p p r e s s o r ( D i o n e x ) w i t h a 0.0125 mol/L H2SO4
rgnrant f l o w i n g a t 2.8 mL/min. Under t h e s e c o n d i t i o n s , the
background conductance was a p p r o x i m a t e l y 16 pS/cm. Most of t h i s
paper i s r e l e v a n t t o o t h e r i o n chromatography c o n f i g u r a t i o n s . In
some ways, t h i s paper i s r e l e v a n t t o a l l i n s t r u m e n t s t h a t d e t e c t a
peak on a n o i s y b a s e l i n e . However, i o n chromatography d i f f e r s
s u b s t a n t i a l l y from gamma s p e c t r o s c o p y and o t h e r measurement
t e c h n i q u e s based on r a d i o a c t i v e decay because the randomness of
r a d i o a c t i v e decay has no a n a l o g i n i o n chromatography.
An i s s u e of c o n s i d e r a b l e importance i n the d i s c u s s i o n of
d e t e c t i o n l i m i t s i s the c h o i c e of s o f t w a r e f o r peak i d e n t i f i c a t i o n
and i n t e g r a t i o n . I n p r a c t i c e , the c h o i c e of a d a t a a n a l y s i s
a l g o r i t h m can have as l a r g e an e f f e c t on the d e t e c t i o n l i m i t as the
c h o i c e o f i n s t r u m e n t c o n f i g u r a t i o n . The d a t a a n a l y s i s i n t h i s paper
has been done w i t h g e n e r a l purpose s t a t i s t i c a l s o f t w a r e r a t h e r t h a n
w i t h one of the p r o p r i e t a r y packages a v a i l a b l e f o r i o n
chromatography. G e n e r a l purpose s o f t w a r e has the advantage of
a l l o w i n g f l e x i b i l i t y i n the d a t a a n a l y s i s . A l s o , g e n e r a l purpose
s o f t w a r e i s based on a l g o r i t h m s t h a t are known p r e c i s e l y .
U n f o r t u n a t e l y , p r o p r i e t a r y packages o f t e n do not come w i t h an e x a c t
s p e c i f i c a t i o n of the a l g o r i t h m s employed. T h i s i s troublesome i n
work on d e t e c t i o n l i m i t s because, as i l l u s t r a t e d below, d e t e c t i o n
l i m i t s can be v e r y s e n s i t i v e t o the c h o i c e of a l g o r i t h m . Of c o u r s e ,
the g e n e r a l purpose s o f t w a r e used i n t h i s paper i s not as c o n v e n i e n t
f o r r o u t i n e l a b o r a t o r y use, as f u l l y d e v e l o p e d , or as w e l l t e s t e d as
the p r o p r i e t a r y packages a v a i l a b l e .
The purpose of t h i s paper i s t o demonstrate a method f o r
e x p l o r i n g l o w - c o n c e n t r a t i o n performance and t o i l l u s t r a t e the d a t a
c h a r a c t e r i s t i c s t h a t such a method t y p i c a l l y r e v e a l s . A d a p t a t i o n s o f
t h e method can be used by any l a b o r a t o r y f o r the e x p l o r a t o r y a n a l y s i s

of chromatograms. There are, however, chromatograms that require

some sophistication i n data analysis. Exploratory methods such as
those discussed i n t h i s paper can be misleading. This paper does not
discuss a l l the ways that the method might be misleading nor a l l the
ways that the method might be improved. In the next section, some of
the needed data analysis steps f o r the analysis of a single
chromatogram are presented. In the t h i r d section, three properties
of chromatograms, the pump cycle, the underlying white noise, and a
low frequency component of unknown o r i g i n are discussed. In the
fourth section, the question of how assessment of the detection l i m i t
i s influenced by t h i s low frequency component and by the choice of
peak i d e n t i f i c a t i o n and integration algorithm i s considered.
I n i t i a l Processing of the Chromatogram
The detection l i m i t f o r p a r t i c u l a analyt i determined b th

peak i n the chromatogra
that obscure t h i s peak.
components that must be removed before the variations that determine
the detection l i m i t can be analyzed. F i r s t , a chromatogram often has
a water dip, some large peaks due to other analytes, and other gross
v a r i a t i o n s . Second, a chromatogram often exhibits a slowly varying
baseline. Third, a chromatogram from a system with a pump might
exhibit a c y c l i c v a r i a t i o n due to the pump. In the chromatograms we
consider, these components can e a s i l y be distinguished from the peak
of interest. The water dip and large peaks due to other analytes do
not coincide with the peak of interest; the baseline i s much smoother
than the peak of interest; and the pump cycle repeats regularly i n
time whereas the peak of interest does not. For these reasons, these
components can be removed from the chromatogram with n e g l i g i b l e
e f f e c t on the detection l i m i t . In other words, these components can
be removed i n such a way that the r e s u l t i n g adjusted chromatogram can
be analyzed as i f these components were never present.
To i l l u s t r a t e the removal of these components, we consider a
chromatogram that i s the r e s u l t of a sample consisting of 0.005 mg/L
n i t r a t e and 0.025 mg/L s u l f a t e . The f i r s t 116 seconds of t h i s
chromatogram consists of a short i n t e r v a l before sample i n j e c t i o n ,
the water dip, and a large peak immediately following the water dip
due perhaps to a solvent inadvertently mixed with the sample. Since
the variations i n the f i r s t 116 seconds are so large, we have
excluded them from Figure 1 and from our analysis of t h i s
chromatogram. Otherwise, Figure 1 shows the chromatogram as i t was
produced by the instrument. As i n other figures i n t h i s paper, we
r e t a i n the o r i g i n f o r the time scale that was set by the instrument.
The most obvious feature i s the sulfate peak. What evidence there i s
of n i t r a t e precedes the sulfate peak by about 120 seconds.
To specify our estimates of the baseline and the pump cycle, we
introduce some notation. chromatogram as
delivered by the instrument. The conductance units are nS/cm. Let
[tQ> tQ + - 1] denote the i n t e r v a l selected f o r analysis. For t h i s
chromatogram, we have tp = 117 and = 799. S h i f t i n g the time o r i g i n
to tQ - 1, we obtain
y (l) ( t ) = Y (0) ( T + t - i), t = 1, ...,

In our estimation o f the baseline and pump cycle, we must exclude the
sulfate peak. To specify the computations, we define a weight
function w(t). This weight function includes the cosine-bell
tapering of the ends of the intervals needed to reduce the bias i n
spectral estimation (2). Let the i n t e r v a l to be excluded because i t
has the sulfate peak be denoted by [ t , t ] . We l e t a D
w(t) = 0 if t <a t < t b
= (1 - cos(irt/21))/2 if 1 < t < 20

= (1 - cos(TT(t-t )/21))/2
a if 1 < t - t
a < 20
= (1 - cos(-n(t-t )/21))/2 b if 1 < t - t b < 20
= (1 - cosU(t-T-l)/21))/2 i f 1 < T-t+1 < 20
= 1 otherwise
If we had decided to exclude other segments of the chromatogram, we

would have set w(t) equa
have set w(t) equal to
segments. The choice o poin tape
another choice might be better.
To remove the baseline, we f i r s t f i t the baseline with a cubic
spline (3). To specify the spline, we choose - 2 i n t e r i o r knots,
which we denote by t , t , 2 t _ i . Let 1 = t ^ < t ^ ... * t =
3 n 2 n
T. In the example being considered, we chose two i n t e r i o r knots, t 2
= 173 and t = 484. These knots divide the i n t e r v a l [1, T] into

3
approximately equal segments. A cubic spline can be characterized as

a function with a continuous second derivative that i s a cubic
polynomial i n each segment ( t j , t j ^ ) . We f i t the cubic spline by
+
l i n e a r least squares using a l l the points i n the chromatogram for

which w(t) 0. Let the f i t t e d baseline be denoted by b ( t ) . The
chromatogram with baseline removed i s given by
y(2)( )t = y<D(t) " b(t)
Because the knots are f a r apart, the f i t t e d baseline i s so smooth

that removing the f i t t e d baseline has no e f f e c t on our a b i l i t y to
distinguish r e l a t i v e l y narrow peaks from other similar variations.
If Figure 1 were plotted again with an expanded time axis, the
periodic v a r i a t i o n due to the pump would be v i s i b l e . To remove the
pump cycle, we f i r s t estimate i t s fundamental frequency from the
spectrum of the chromatogram. The Fourier transform of the
chromatogram that we use ^n our estimation i s actually the Fourier
in our x
x
transform of w(t) times y

2
z(f) = w(t)y( )(t)exp(-i2-fff(t-l))
t=l
Note that z ( f ) equals the complex conjugate of z ( l - f ) and that z ( f )

equals z ( f - n ) , where i s an integer. Thus, we can compute z ( f ) for
0 < f < .5 and obtain i t s values for a l l f. Using the fast Fourier
transform algorithm, we computed z ( f ) for f = j/N, where = 6720 and
j = 0, 3360. We chose a value of much larger than so that
we could estimate the fundamental of the pump cycle with s u f f i c i e n t
accuracy. The fundamental frequency of the pump cycle, which we
denote by fp, can be estimated from the peaks of the spectral

2
estimate | z ( f ) | . There i s r e a s o n t o be concerned about t h e
p r o p e r t i e s o f t h i s s p e c t r a l e s t i m a t e because o f t h e gap i n t h e s e r i e s
where t h e s u l f a t e peak was. S p e c t r a l e s t i m a t i o n f o r s e r i e s w i t h
m i s s i n g d a t a have been d i s c u s s e d ( 4 ) . F o r o u r example, we o b t a i n e d
f g = 1129/6720. T h i s f r e q u e n c y c o r r e s p o n d s t o a p e r i o d o f
a p p r o x i m a t e l y 6 seconds.
Our c o r r e c t i o n f o r t h e pump c y c l e i s based on t h e p r e m i s e t h a t

t h e pump c y c l e i s p e r f e c t l y s t a b l e o v e r t h e e n t i r e chromatogram,
a l t h o u g h , as shown below, t h i s i s n o t always t r u e . Based on t h e
spectrum o f t h e chromatogram, we have c o n c l u d e d t h a t t h e fundamental
and t h e f i r s t f o u r harmonics c o n t a i n v i r t u a l l y a l l o f t h e pump c y c l e .
For t h e s e r e a s o n s , we use as an e s t i m a t e o f t h e pump c y c l e
5
p ( t ) = 2Re[ z ( k f ) e x p ( i 2 * k f ( t - l ) ) ] / [ w(t) ]
k=l t=
I f 2kfQ i s an i n t e g e r , t h e n t h i s f o r m u l a must be a d j u s t e d by
substituting z(kf )/2 forz ( k f ) .
Q 0 We o b t a i n as o u r a d j u s t e d
chromatogram
y(3)( ) = y(2)( ) - p( )
t t t
I f t h e pump c y c l e i s indeed s t a b l e , t h e n t h i s c o r r e c t i o n f o r t h e pump

c y c l e does n o t d i s t o r t peaks o f i n t e r e s t because t h e e s t i m a t e p ( t ) i s
o b t a i n e d from a l o n g r e c o r d , t h e e n t i r e chromatogram.
The a d j u s t e d chromatogram f o r o u r example i s shown i n F i g u r e 2.
The s u l f a t e peak s t a n d s o u t . Some e v i d e n c e o f t h e n i t r a t e peak can
be seen about 120 seconds b e f o r e t h e s u l f a t e peak. Comparison o f
F i g u r e s 1 and 2 shows t h a t removal o f t h e pump c y c l e does reduce what
i n F i g u r e 1 appears t o be random n o i s e .
Models o f Chromatogram N o i s e
The a d j u s t e d chromatogram, w h i c h i s i l l u s t r a t e d i n F i g u r e 2, p r o v i d e s
a s t a r t i n g p o i n t f o r d e t a i l e d modeling o f t h e v a r i a t i o n s i n t h e
chromatogram t h a t i n t e r f e r e most w i t h t h e d e t e c t i o n o f a n a l y t e s a t
low c o n c e n t r a t i o n s . I n t h i s m o d e l i n g , we c o n c e n t r a t e on t h r e e
components, t h e pump c y c l e , v e s t i g e s o f w h i c h may remain because o f
pump c y c l e i n s t a b i l i t y , t h e u n d e r l y i n g w h i t e n o i s e , and some low
f r e q u e n c y v a r i a t i o n s t h a t might be m i s t a k e n f o r peaks o f i n t e r e s t .
A d i r e c t way t o i n v e s t i g a t e t h e s t a b i l i t y o f t h e pump c y c l e i s
t o e s t i m a t e t h e pump c y c l e a t v a r i o u s p o i n t s a l o n g t h e chromatogram.
One approach t o t h i s e s t i m a t i o n i s i n v e r s e F o u r i e r t r a n s f o r m a t i o n o f
t h e F o u r i e r c o e f f i c i e n t s z ( f ) t h a t l i e c l o s e t o t h e pump c y c l e
fundamental and i t s harmonics. T h i s approach i s a v e r s i o n o f complex
d e m o d u l a t i o n ( 5 ) . L e t M - 1 be t h e number o f harmonics t o be
i n c l u d e d ; and l e t
z ( f ) = z ( f ) i f |fN - m o d ( k f N , N ) | < 24 o r |f - mod(N-kfQN,N)| < 24

M 0
f o r k = 1 o r 2 o r ... o r M
= 0 otherwise
An e s t i m a t e o f t h e pump c y c l e t h a t a l l o w s some v a r i a t i o n over t h e

chromatogram i s g i v e n by

Conductance (nS/cm)
Tine (seconds)
F i g u r e 1. Chromatogram produced by. t h e i n s t r u m e n t .
Adjusted Conductance (nS/cn)
F i g u r e 2. Chromatogram i n i t i a l l y a d j u s t e d f o r b a s e l i n e and pump

cycle.

N-l
p(t) = (1/N) z (j/N)exp(i2irj(t-l)/N)
5
j=0
Figure 3 shows the pump cycle for a chromatogram d i f f e r e n t from

the one considered i n the previous section. The fundamental of the
pump cycle i n t h i s case i s 1128/6720. The shape of the pump cycle i s
shown every 50 periods, which i s approximately every 300 seconds. In
the figure, successive traces are o f f s e t by 1 nS/cm. The shapes
shown are t y p i c a l of those produced by the two piston pump that we
used. This chromatogram shows considerable i n s t a b i l i t y . The pump
cycle i n the chromatogram considered i n the previous section i s much
more stable as i s the pump cycle i n another chromatogram that we
consider i n d e t a i l below. I f the pump cycle i s unstable, then an
estimate such as p(t) may be more appropriate for the removal of the
pump cycle than the estimat
The dependence of
instrument i s of interest. As noted above, the chromatograph i s
generally operated with a flow rate of 2.0 mL/min. Reduction of the
flow rate i s accomplished by lengthening the period between strokes.
The reduction to 1.5 mL/min lengthens the period from approximately 6
seconds to approximately 8 seconds.
Variations i n the strength of the pump cycle are also of
interest. Because the pump cycle i s p o t e n t i a l l y unstable, we
consider the power i n frequency bands about the fundamental and the
f i r s t two harmonics instead of the amplitude of the pump cycle as i n
the previous section. Our index of pump cycle power i s given by
N-l
2 2
[ |z (j/N)|
3 ]/[ K 3 w (t) ]
j-1 t-1
where K i s the number of values of j i n the sum for which z ( j / N )

3 3
0. The pump cycle power can be quite variable. The chromatogram

discussed i n the previous section has a pump cycle power of 6.6. Two
other chromatograms obtained under the same conditions showed pump
cycle powers of 5.8 and 12.3.
A series of chromatograms were obtained to investigate the
dependence of the pump cycle power on the eluent. These
chromatograms were obtained with the guard column AG4A (Dionex) i n
place. The f i r s t and l a s t chromatograms, which were run with the
standard eluent, gave pump cycle powers of 5.9 and 2.6. The second
chromatogram, which was run with a weaker eluent 0.75 mmol/L Na C0 , 2 3
gave a pump cycle power of 4.0. This eluent gives a background

conductance of 9.6 pS/cm instead of the standard 16 pS/cm. The t h i r d
chromatogram, which was run with deionized water, gave a pump cycle
power of 0.1. The background conductance of deionized water i s
approximately 2.6 pS/cm. This t h i r d chromatogram provides support
for the hypothesis that the pump cycle power i s lower for weaker
eluents. Additional research i s planned to investigate these aspects
more f u l l y .
A random noise component with f l a t spectrum i s evident i n the
adjusted chromatograms. This component seems to be caused by the

11. KOCH & LIGGETT Critical Assessment of Detection Limits 21T
electronics of the instrument because i t s standard deviation does not

vary with chromatographic conditions. The standard deviation of t h i s
underlying white noise i s approximately 0.4 nS/cm. This value can be
compared to the size of the pump cycles shown i n Figure 3.
The noise shown i n Figure 2 does not have a f l a t spectrum,
however. The power spectrum i s higher at low frequencies. To
display the low frequency part of the adjusted chromatogram, we
smooth the adjusted chromatogram using a f i l t e r with impulse response
h ^ t ) = (1 + cos(iTt/8))/16 i f -7 < t < 7

= 0 otherwise
This cosine b e l l i s 8 seconds wide at half height and thus seems to

match the peak width expected of a small n i t r a t e peak. Figure 4
shows the smoothed adjusted chromatogram given by
I h (j) (t-j)y(3)(t-j)
1 W
j=-7
Note that we have included the weight function w(t) to suppress the
s u l f a t e peak. Figure 4 has two equal height peaks on the l e f t side.
The right most of these i s the n i t r a t e peak. This peak can be
compared to the other peaks i n t h i s smoothed chromatogram. Clearly,
there are several peaks that might be mistaken for a peak due to an
analyte of interest.
Figure 5 shows another chromatogram that we have adjusted and
smoothed i n the same way as Figure 4. This chromatogram i s the f i r s t
of the series we obtained to investigate the dependence of the pump
cycle on the eluent. No sample was injected i n the generation of
t h i s chromatogram. Nevertheless, t h i s chromatogram also shows
several peaks that might be mistaken for analyte.
The cause of the low frequency noise shown i n Figures 4 and 5 i s
an important question. In most applications of ion chromatography,
two p o s s i b i l i t i e s can be suggested. One i s sample-to-sample
carryover due either to contamination i n the sample i n j e c t i o n loop or
to a slowly eluting organic l e f t on the separator column by a
previous sample. The other i s mechanical transients. The existence
of the pump cycle suggests that variations i n flow past the
conductivity detector cause variations i n the chromatogram.
Mechanical transients can also cause variations i n the flow. Other
sources of noise are also possible, and w i l l be investigated i n
future research.
Can t h i s low frequency component be treated as though i t were
generated by a random mechanism so that a s t a t i s t i c a l statement can
be made about i t ? Sample-to-sample carryover can be treated
s t a t i s t i c a l l y only under very special conditions on the order i n
which the samples are analyzed. Mechanical transients can be treated
s t a t i s t i c a l l y only i f t h e i r source i s random i n some sense. The best
solution to t h i s low frequency component i s to reduce i t s s i z e u n t i l
the question of i t s randomness i s no longer important. This might be
done by reducing i t s magnitude or a l t e r n a t i v e l y by i n j e c t i n g each
sample twice or more. In any case, any laboratory that does
extensive low concentration ion chromatography analysis should

PUMP Cycle (nS/ciO
-1
0 Time within Period (seconds) 6
F i g u r e 3. Shape o f t h e pump c y c l e e v e r y 50 p e r i o d s .
Smoothed Conductance (nS/cn)
: : /. '. A $
: : S : ' : . . /;
* \ ,':. V \ :' V ; :
'' : V V
<
' 'V
i
180. Tine (seconds) 1000.
F i g u r e 4. Chromatogram i n i t i a l l y a d j u s t e d and smoothed, example

w i t h n i t r a t e peak.

experiment w i t h t h i s component t o see what type o f s t a t i s t i c a l

inferences are v a l i d .
Detection Limits
Under t h e a s s u m p t i o n t h a t t h e low frequency component i s a s t a t i o n a r y

G a u s s i a n random p r o c e s s , l e t us proceed w i t h t h e s p e c i f i c a t i o n o f a
d e t e c t i o n a l g o r i t h m . The d e t e c t i o n l i m i t i n d i c a t e s t h e performance
o f t h e d e t e c t i o n a l g o r i t h m . D e t e c t i o n o f peaks o f i n t e r e s t i n t h e
a d j u s t e d chromatogram i n v o l v e s removal o f t h e low f r e q u e n c y component
t o t h e e x t e n t p o s s i b l e and perhaps t h e s e a r c h i n t i m e f o r t h e d e s i r e d
peak.
F i g u r e s 4 and 5 show v a r i a t i o n s w i t h t i m e t h a t a r e much more
g r a d u a l than t h e peaks o f i n t e r e s t . These v a r i a t i o n s can be removed
i n v a r i o u s ways. C o n s i d e r f i r s t t h e use o f a f i l t e r w i t h t h e
f o l l o w i n g impulse respons
h ( t ) = h ( t ) - (1
2 x
= 0 otherwise.
T h i s f i l t e r i s t h e d i f f e r e n c e between t h e smoother used t o o b t a i n

F i g u r e s 4 and 5 and a broader c o s i n e - b e l l smoother. A p p l y i n g t h i s
f i l t e r t o t h e a d j u s t e d chromatogram d i s c u s s e d i n t h e second s e c t i o n ,
we o b t a i n t h e f i l t e r e d chromatogram shown i n F i g u r e 6. Note t h a t
s i n c e we have suppressed t h e s u l f a t e peak as we d i d i n F i g u r e 4, t h e
n i t r a t e peak i s now t h e l a r g e s t peak. The peaks i n F i g u r e 6 can be
compared t o a t h r e s h o l d a p p r o p r i a t e t o t h e u n d e r l y i n g w h i t e n o i s e
component. The f i l t e r i n g changes t h e s t a n d a r d d e v i a t i o n o f t h i s
component from 0.4 t o 0.064. Thus, t h e a p p r o p r i a t e t h r e s h o l d f o r a
s i g n i f i c a n c e l e v e l o f 0.05 i s 1.645-0.064 = 0.11. F i g u r e 6 s u g g e s t s
t h a t t h i s t h r e s h o l d i s t o o low and t h a t t h e f i l t e r i n g has n o t removed
t h e low f r e q u e n c y component e n t i r e l y . N e v e r t h e l e s s , t h i s t h r e s h o l d
i s n e a r l y r i g h t , and t h e u n d e r l y i n g w h i t e n o i s e seems t o be t h e major
c o n t r i b u t o r t o t h e f i l t e r e d chromatogram.
Another way t o remove t h e low f r e q u e n c y component i s t o f i t a
s p l i n e t h a t i s n o t as smooth as t h e s p l i n e f i t i n t h e second s e c t i o n .
We f i t a q u a d r a t i c s p l i n e (one t h a t i s a q u a d r a t i c p o l y n o m i a l between
k n o t s and t h a t has a c o n t i n u o u s f i r s t d e r i v a t i v e ) . We choose knots
e v e r y 50 seconds and f i t t h e s p l i n e by l e a s t squares t o t h e a d j u s t e d
chromatogram. C o n s i d e r t h e a d j u s t e d chromatogram t h a t was smoothed
and shown i n F i g u r e 5. A f t e r s u b t r a c t i n g t h e f i t t e d q u a d r a t i c s p l i n e
from t h i s a d j u s t e d chromatogram, t h e r e b y f u r t h e r a d j u s t i n g i t , and
then smoothing t h e r e s u l t w i t h t h e f i l t e r h - ^ ( t ) , we o b t a i n t h e
smoothed, a d j u s t e d chromatogram shown i n F i g u r e 7. The t h r e s h o l d f o r
t h i s f i g u r e based on t h e u n d e r l y i n g w h i t e n o i s e i s 0.20. Once a g a i n ,
t h e t h r e s h o l d seems t o be t o o low but seems t o be c l o s e enough t o
suggest t h a t t h e u n d e r l y i n g w h i t e n o i s e i s t h e dominant c o n t r i b u t o r
t o t h e smoothed, a d j u s t e d chromatogram shown.
Most s o f t w a r e f o r i o n chromatography i n c l u d e s t h e o p t i o n t o
s e a r c h i n t i m e f o r t h e peak o f i n t e r e s t . C o n s i d e r t h e f o l l o w i n g
s i m p l e a l g o r i t h m f o r f i n d i n g t h e n e a r e s t peak:
1. I n i t i a l i z e l e f t maximum and r i g h t maximum w i t h t h e v a l u e a t t h e

t i m e where t h e peak i s expected.
2. S e a r c h l e f t f o r a h i g h e r maximum u n t i l a v a l u e i s encountered

Snoothed Conductance (nS/cn)
m IS. Ha
i ^
9 TIM (seconds) 2888.
F i g u r e 5. Chromatogram i n i t i a l l y a d j u s t e d and smoothed, example

w i t h o u t n i t r a t e peak.
Filtered Adjusted Conductance (nS/cit)

,5

' ' ! ' . . :
: , '' \ !
v ! ..'V' ;'. '.
'iv ' :: ':
.5.
188. Tine (seconds) 1888.
F i g u r e 6. Chromatogram i n i t i a l l y a d j u s t e d and f i l t e r e d , example

w i t h n i t r a t e peak.

that i s less than the current l e f t maximum by a specified

amount.
3. Search right for a higher maximum u n t i l a value i s encountered
that i s less than the current right maximum by a specified
amount.
4. Choose the larger of the l e f t and right maxima.
We have applied t h i s algorithm to the smoothed, adjusted

chromatogram shown i n Figure 7 using as starting points a sequence of
times that are 30 seconds apart. The specified amount used to end
the search was 0.07. We can compare the d i s t r i b u t i o n of the results
of t h i s to the d i s t r i b u t i o n of a l l the points shown i n Figure 7. We
do t h i s by p l o t t i n g the quantiles of the d i s t r i b u t i o n of nearest
peaks against the corresponding quantiles of the d i s t r i b u t i o n of a l l
the points (6). The result i s shown i n Figure 8. I f the two
d i s t r i b u t i o n s were the same then the points would l i e along the 45
degree l i n e through th
o f f s e t with respect to th
which i s not surprising. Interestingly, the slope of the points i s
close to 1 suggesting that the variances of the two d i s t r i b u t i o n s are
nearly the same.
Figure 8 provides one i l l u s t r a t i o n of the difference between a
detection algorithm that searches and one that does not. With an
algorithm that does not search, any point i n Figure 7 might be the
noise contribution to the observed analyte peak. With an algorithm
that searches, the d i s t r i b u t i o n of the noise contribution depends on
the concentration of the analyte. With no analyte, the d i s t r i b u t i o n
of the noise contribution i s the d i s t r i b u t i o n of noise peaks.
However, the presence of analyte influences the search. With
s u f f i c i e n t analyte, the analyte determines the r e s u l t of the search.
For t h i s reason, i n the case of an algorithm that searches, the
d i s t r i b u t i o n of the noise contribution to the observed analyte peak
can be thought of as lying between the two d i s t r i b u t i o n s compared i n
Figure 8. This fact complicates the s t a t i s t i c a l inference for
measurements near the detection l i m i t . I t also suggests that the
determination of detection l i m i t s from measurements on the blank may
not be a v a l i d procedure when the software employed searches for the
peak.
If no search for the n i t r a t e peak i s needed, i f the low
frequency component can be treated as stationary and Gaussian, and i f
the contribution to the chromatogram from the analyte does not vary,
then the results i n t h i s paper provide an assessment of the n i t r a t e
detection l i m i t . Figure 6 shows that 0.005 mg/L i s close to the
detection l i m i t . I f we add to the standard deviation of the
underlying white noise an amount to compensate for the low frequency
component and multiply by the usual factor 3.29, we obtain a
comparable number. The amount to be added can be judged from Figures
6 and 7. Of the three conditions on the v a l i d i t y of t h i s assessment,
the f i r s t two are discussed above. The t h i r d , the condition on the
contribution from the analyte, i s also important. V a r i a t i o n i n the
contribution from the analyte might be caused by the separator
column, for example. Several chromatograms with r e p l i c a t e sample
injections with the same concentration of n i t r a t e are needed to
assess t h i s v a r i a t i o n . This w i l l form the basis of the second phase
of t h i s research i n the evaluation of detection l i m i t s .

Smoothed Adjusted Conductance (nS/cM)

.6"
:
" ^ >. .. " !"
:
':'.*.'.-'-'
: - ! > .V ; {
.5.
TiHe (seconds) 288.
F i g u r e 7. Chromatogram i n i t i a l l y a d j u s t e d , f i n e l y a d j u s t e d f o r
b a s e l i n e , and smoothed, example w i t h o u t n i t r a t e peak.
Quantiles of the Set of Nearest Peaks
F i g u r e 8. Q u a n t i l e - q u a n t i l e p l o t o f n e a r e s t peaks and a l l
points.

Disclaimer
C e r t a i n commercial equipment, i n s t r u m e n t s , o r m a t e r i a l s a r e
i d e n t i f i e d i n t h i s paper i n o r d e r t o s p e c i f y a d e q u a t e l y t h e
e x p e r i m e n t a l procedure. Such i d e n t i f i c a t i o n does n o t i m p l y
recommendation o r endorsement by t h e N a t i o n a l Bureau o f S t a n d a r d s ,
nor does i t i m p l y t h a t t h e m a t e r i a l s o r equipment i d e n t i f i e d a r e
n e c e s s a r i l y t h e best a v a i l a b l e f o r t h e purpose.
Literature Cited
1. Wetzel, R. .; Pohl, C. .; Riviello, J . M.; MacDonald, J. C.
In Inorganic Chromatographic Analysis; MacDonald, J. C., Ed.;
John Wiley and Sons: New York, 1985; p 355.
2. Tukey, J . W. In Spectral Analysis of Time Series; Harris, .,
Ed.; John Wiley and Sons: New York 1967; p 25
3. De Boor, C. A Practica
York, 1978; Chapte
4. Marquardt, D. W.; Acuff, S. K. In Applied Time Series Analysis;
Anderson, O. D.; Perryman, M. R., Eds.; North Holland Publishing
Co.: Amsterdam, 1982; p 199.
5. Bloomfield, P. Fourier Analysis of Time Series: An
Introduction; John Wiley and Sons: New York, 1976; Chapter 6.
6. Chambers, J . M.; Cleveland, W. S.; Kleiner, B . ; Tukey, P. A.
Graphical Methods for Data Analysis; Wadsworth International
Group: Belmont, California, 1983; Chapter 3.

Chapter 12
Risk Assessment as a Tool

for Verifiable Detection and Quantification
of Fusarium Trichothecenes in Human Blood
at Low Parts-Per-Billion Concentrations
D. J. Reutter, S. F. Hallowell, and E. W. Sarver
Research Directorate, Development and Engineering Center, U.S. Army

Chemical Research, Aberdeen Proving Ground, MD 21010-5423
A n a l y t i c a l chemists are asked often to determine the

presence of a chemical in a s p e c i f i c matrix at some
stated concentration. Examples of such a n a l y t i c a l
procedures include the determination of c o n t r o l l e d
substances and toxins i n the b i o l o g i c a l f l u i d s of victims
of alleged poisonings. Reports of the a n a l y t i c a l findings
in these cases frequently become the object of legal
decisions or international p o l i c i e s . The d e f i n i t i o n s of
detection l i m i t s in these instance can be s i g n i f i c a n t l y
different from those commonly used i n research and
industrial laboratories. Stringent q u a l i t y assurance and
q u a l i t y control throughout the e n t i r e a n a l y t i c a l process
is required to e s t a b l i s h and maintain statistically
definable l i m i t s of detection and q u a n t i f i c a t i o n and the
c e r t a i n t y of i d e n t i f i c a t i o n . The q u a l i t y assurance plan
developed and employed by t h i s Center for the
determination and q u a n t i f i c a t i o n of Fusarium mycotoxins
in human blood i s reported. Under t h i s plan, the
a n a l y t i c a l method i s divided into four steps; (1) sample
and standard handling, (2) sample preparation ana clean-
up, (3) d e r i v a t i z a t i o n and (4) a n a l y s i s . Each step i s
evaluated to ascertain where errors occur, and s p e c i f i c
q u a l i t y control procedures are introduced i n each step
to detect, i s o l a t e and correct errors during the a n a l y s i s .
Limits of detection, v e r i f i c a t i o n and q u a n t i f i c a t i o n are
i n d i v i d u a l l y determined and v a l i d a t e d .
R e c e n t l y , few t o p i c s i n a n a l y t i c a l c h e m i s t r y h a v e o c c u p i e d t h e
s c i e n t i f i c c o m m u n i t y more t h a n t h e a b i l i t y o f c h e m i c a l l a b o r a t o r i e s
t o r e l i a b l y d e t e r m i n e a t t h e low p a r t s - p e r - b i l l i o n l e v e l t h e
p r e s e n c e o f F u s a r i u m t r i c h o t h e c e n e s i n e n v i r o n m e n t a l and
t o x i c o l o g i c a l samples. T h i s paper provides a s y s t e m a t i c approach
f o r d e v e l o p i n g and i m p l e m e n t i n g a q u a l i t y a s s u r a n c e and q u a l i t y
c o n t r o l p r o g r a m f o r a c o m p l e x a n a l y t i c a l m e t h o d i n w h i c h human e r r o r
and s y s t e m f a i l u r e c a n o c c u r . The a p p l i c a t i o n o f t h i s approach
to t h e problem o f determining the presence o f nine naturally


12. REUTTER ET AL. Risk Assessment as a Tool 225
o c c u r r i n g m y c o t o x i n s and t h e i r m e t a b o l i t e s i n human b l o o d a t low

p a r t - p e r - b i l l i o n (ppb) c o n c e n t r a t i o n s i s g i v e n . The key t o a
s u c c e s s f u l program i s the development o f a r i s k a n a l y s i s f o r the
a n a l y t i c a l procedure which e x p l i c i t l y r e c o g n i z e s the p o t e n t i a l f o r
human e r r o r . A n e c e s s a r y s t e p i n d e v e l o p i n g an a n a l y t i c a l s t r a t e g y
i s t h e r e - s t a t e m e n t by t h e a n a l y s t o f p r o g r a m r e q u i r e m e n t s n e c e s s a r y
t o d e l i n e a t e t h e i n t e n d e d use o f d a t a and t h e s i g n i f i c a n c e o f
r e p o r t i n g f a l s e p o s i t i v e , f a l s e n e g a t i v e , and i m p r e c i s e r e s u l t s .
I t i s o n l y a f t e r t h e s e r e q u i r e m e n t s a r e d e f i n e d , and a r e l i a b l e
a n a l y t i c a l method i s d e v e l o p e d , t h a t d e t e c t i o n l i m i t s w h i c h a r e
a p p r o p r i a t e f o r t h e a n a l y s i s can be c a l c u l a t e d . T h e n , s p e c i f i c
q u a l i t y c o n t r o l m e a s u r e s t h a t e l i m i n a t e o r q u a n t i f y e r r o r s must be
d e v e l o p e d and i m p l e m e n t e d . The i m p l e m e n t a t i o n o f s u c h a p l a n f o r
t h e a n a l y s i s o f a s e t o f samples p r e p a r e d by an i n d e p e n d e n t
l a b o r a t o r y and a n a l y z e d b l i n d l y by t h i s l a b o r a t o r y f o l l o w s .
Trichothecene mycotoxins are secondary metabolites o f v a r i o u s fungal
s p e c i e s . S t r u c t u r e s o f som
t h e US ARMY a r e g i v e n i
r e p o r t e d f o r t h e a n a l y s i s o f t h e s e t o x i n s (1-11, 1 5 ) . Of t h e s e , mass
s p e c t r o m e t r y t e c h n i q u e s a r e b o t h s e n s i t i v e and d e f i n i t i v e when
a p p l i e d t o t o x i c o l o g i c and e n v i r o n m e n t a l s a m p l e s . W i t h c u r r e n t
t e c h n o l o g y , t h e most s e n s i t i v e and q u a l i t a t i v e l y d e f i n i t i v e
a n a l y t i c a l technique f o r the determination of these t o x i n s i s
d e r i v a t i z a t i o n w i t h an e l e c t r o n d e f i c i e n t m o i e t y f o l l o w e d by
a n a l y s i s w i t h n e g a t i v e i o n c h e m i c a l i o n i z a t i o n gas chromatography-
mass s p e c t r o m e t r y (NICI-GC/MS).
The a n a l y t i c a l p r o c e d u r e t h a t i s u s e d by t h i s l a b o r a t o r y f o r t h e
a n a l y s i s o f s i m p l e F u s a r i u m m y c o t o x i n s w i l l be r e p o r t e d s e p a r a t e l y .
However, t h e a n a l y t i c a l scheme i s o u t l i n e d i n F i g u r e 2. The
method i s v e r y a r d u o u s due t o s e v e r a l s a m p l e c l e a n - u p s t e p s w h i c h
n e c e s s i t a t e s t r a n s f e r o f t h e s a m p l e between c o n t a i n e r s . The
t r i c h o t h e c e n e s and t h e i r d e r i v a t i v e s h a v e a t e n d e n c y t o a d h e r e t o
g l a s s and c a n be q u a n t i t a t i v e l y t r a n s f e r r e d o n l y w i t h numerous
methanol w a s h e s . W h i l e t h e a n a l y t i c a l method i s b o t h s u f f i c i e n t l y
s e n s i t i v e and d e f i n i t i v e f o r t h e p r o g r a m r e q u i r e m e n t s , t h e s h e e r
amount o f human m a n i p u l a t i o n r e q u i r e d f o r t h e c o m p l e t i o n o f t h i s
a n a l y s i s makes i t somewhat u n r e l i a b l e i f i m p l e m e n t e d w i t h o u t a
r e s p o n s i b l e q u a l i t y a s s u r a n c e and q u a l i t y c o n t r o l p r o g r a m .
In an e m e r g i n g and c o m p l e x a n a l y t i c a l m e t h o d o l o g y , t h r e e f a c t o r s
must a l w a y s be c o n s i d e r e d . F i r s t , as a l w a y s , t h e d e t e c t i o n c r i t e r i a
must be e s t a b l i s h e d a p r i o r i . , a l o n g w i t h t h e c r i t e r i a f o r t h e r i s k
the l a b o r a t o r y w i l l accept i n r e p o r t i n g the r e s u l t s . This includes
d e t e r m i n a t i o n o f t h e p r o b a b l e r a t e o f r e p o r t i n g f a l s e p o s i t i v e s and
f a l s e n e g a t i v e s f o r the a n a l y s i s . Second, a f u l l r i s k assessment
must be c o n d u c t e d t o e x p l o r e on p a p e r , even b e f o r e e n t e r i n g t h e
l a b o r a t o r y , where c a t a s t r o p h i c f a i l u r e s i n t h e a n a l y t i c a l
methodology might occur. T h i r d , i n view o f the r i s k assessment, a
q u a l i t y a s s u r a n c e / q u a l i t y c o n t r o l program which w i l l prevent
c a t a s t r o p h i c f a i l u r e s and m e a s u r e t h e p e r f o r m a n c e o f t h e a n a l y s i s
must be i m p l e m e n t e d .
DETECTION, V E R I F I C A T I O N AND QUANTIFICATION C R I T E R I A

The d e v e l o p m e n t o f d e t e c t i o n c r i t e r i a a l w a y s r e s u l t s f r o m p r o g r a m
r e q u i r e m e n t s i n i t i a t e d by some g e n e r a l management g u i d e l i n e . In

T-2 TYPE
. H H H
R
B 2
"3
Trichothecene R R
2 3 4
Basic Trichothecene H H H H
Trichodermol (roridin C) Fi OH H H H
Trichodermin H OAc a
H H H
* Verrucarol H OH OH H H
* Scirpentrioi
Monoucetoxyscirpenol (MAS
* Diucetoxyscirpenol (anguidineKDAS) OH OAc OAc H H
7-Hydroxy DAS OH OAc OAc OH H
Calonectrin OAc H OAc H H
15-Diacctylcalonectrin OAc H OH H H
Dihydroxy trichothecene H OH H H OH
T-2 tetraol OH OH OH H OH
Neosolaniol (solaniol) OH OAc OAc H OH
Monoacetylneosolaniol OH OAc OAc H OAc
7.8-Dihydroxy DAS OH OAc OAc OH OH
* HT-2 toxin OH OH OAc H b
* T-2 toxin OH OAc OAc H b
Acetyl T-2 toxin OAc OAc OAc H b
- Toxins used in Study

- 0-C-CH-j; Acetate
0
b - 0 - C - C H - C H ( c : H ) 2 : Isovulerate
2 3
NIVALENOL TYPE
Trichothecene R, R R R
:
3 4
Nivulenol OH OH OH OH
Monoacetylnivaienol (fusarenon-X) OH OAc OH OH
Diucetyinivaienol (DAN) OH OAc OAc OH
* Deoxynivalenol (DON) (vomitoxin) OH OH OH
Monoacetyl DON OAc OH OH
Diacetyi DON OAc OAc OH
Trichothecin a
Trichothecolone OH
F i g u r e 1. S t r u c t u r e o f Some Simple T r i c h o t h e c e n e M y c o t o x i n s

I RECEIPT OF SAMPLE I
LOG-ASSIGNMENT OF NUMBER
STORAGE FROZEN
1
REMOVAL OF SAMPLE
FROM STORAGE
i ~
CENTRIFUGATION

ETHYL ACETATE EXTRACTION
OF LIPOPHILES
EXTRACTS
* HYDROPHILES
L-CHROMATOGRAPHIC
CLEAN-UP MeOH/ACETONE PRECIPITATION
(NORMAL PHASE)
TRANSFERRAL OF SAMPLES L.C. CLEAN-UP

TO Rxn VIALS (REVERSE PHASE)
DERIVATIZATION WITH
APPROPRIATE REAGENT BLOW TO DRYNESS
-_t *
INSTRUMENTAL
ANALYSIS
*
STATISTICAL EVALUATION
DATA
REPORT
F i g u r e 2. Procedure f o r D e t e r m i n a t i o n o f T r i c h o t h e c e n e s i n Blood

t h i s c a s e , t o x i c o l o g i c a l s t u d i e s have e s t a b l i s h e d t h a t
c o n c e n t r a t i o n s a s l o w a s 10 p p b c a n b e s i g n i f i c a n t f o r some
t r i c h o t h e c e n e m y c o t o x i n s when p r e s e n t i n t h e b l o o d o r t i s s u e s o f
humans o r o t h e r mammals(12-15). T h i s i s t h e d e t e c t i o n l i m i t
c r i t e r i o n w h i c h s e r v e d a s t h e benchmark f o r d e v e l o p m e n t o f t h e
a n a l y t i c a l m e t h o d o l o g y ; however, i t d o e s n o t d e f i n e t h e d e t e c t i o n
l i m i t f o r t h e a n a l y s i s as i t i s p r a c t i c e d and r e p o r t e d o u t o f t h e
l a b o r a t o r y . One o f t h e most i m p o r t a n t t a s k s o f a p r o g r a m manager
i s t o t r a n s l a t e t h e general program requirements i n t o t e c h n i c a l
c r i t e r i a . T h e p r o g r a m r e q u i r e m e n t s f o r t h i s s t u d y were s i m p l y t o
d e v e l o p a n a n a l y t i c a l m e t h o d o l o g y w h i c h was s u f f i c i e n t l y s e n s i t i v e
t o d e t e c t and v a l i d a t e t h e presence o f t r i c h o t h e c e n e s a t
t o x i c o l o g i c a l l y s i g n i f i c a n t l e v e l s with a very high degree o f
c o n f i d e n c e . T h i s was t r a n s l a t e d i n t o t e c h n i c a l r e q u i r e m e n t s a s
f o l l o w s . 1) T h e q u a l i t y o f t h e d a t a i s more i m p o r t a n t t h a n s a m p l e
t h r o u g h p u t . 2 ) F a l s e p o s i t i v e s a r e more d e t r i m e n t a l t h a n f a l s e
n e g a t i v e s however b o t h modes o f f a i l u r e a r e v e r y s e r i o u s . 3 ) Minimum
d e t e c t a b l e l i m i t s a n d minimu
same r a n g e o r l o w e r t h a
s i g n i f i c a n c e . 4) Q u a l i t y c o n t r o l must b e r u n a t a l e v e l w h i c h w i l l
demonstrate t h e q u a n t i t a t i v e p r e c i s i o n o f t h e a n a l y s i s . 5)
Q u a l i t a t i v e c e r t a i n t y i s more i m p o r t a n t t h a n q u a n t i t a t i v e a c c u r a c y .
The s p e c i f i c c r i t e r i a w h i c h w e r e d e v e l o p e d t o d e f i n e minimum
d e t e c t i o n l i m i t , v e r i f i c a t i o n l i m i t and range o f q u a n t i f i c a t i o n a r e
given i n Table I .
EVALUATION OF S E L E C T I V I T Y FOR AN ANALYTICAL PROCEDURE

Due t o p o l i t i c a l a n d i n t e r n a t i o n a l i m p l i c a t i o n s o f r e p o r t i n g a f a l s e
p o s i t i v e r e s u l t , t h i s t y p e o f e r r o r w e r e c o n s i d e r e d t h e most s e r i o u s
e r r o r p o s s i b l e i n t h i s a n a l y s i s . Therefore, t h e a n a l y t i c a l procedure
was e x a m i n e d t o d e t e r m i n e t h e p o t e n t i a l f o r t h i s t y p e o f e r r o r
t o o c c u r . T h e b e s t method o f d o i n g t h i s i s b y e x p e r i m e n t a l l y
a n a l y z i n g a l a r g e number o f s a m p l e s w h i c h i n c l u d e a s i g n i f i c a n t
number i n w h i c h t h e a n a l y t e i s known t o b e b e l o w t h e d e t e c t i o n
l i m i t o f the procedure. In p r a c t i c e , t h i s i s seldom done both
because o f t h e c o s t i n v o l v e d i n performing such a study and t h e
probable n o n a v a i l a b i l i t y o f standard samples which a r e well
c h a r a c t e r i z e d a t t h e c o n c e n t r a t i o n l e v e l a t w h i c h a new, v e r y
s e n s i t i v e a n a l y t i c a l t e c h n i q u e i s being a p p l i e d . An a l t e r n a t i v e
to t h i s i s t o examine t h e a n a l y t i c a l procedure t h e o r e t i c a l l y .
U s i n g K a i s e r ' s ( 1 6 ) d e f i n i t i o n f o r c a l c u l a t i n g t h e " i n f o r m i n g power"
o f a n a n a l y t i c a l p r o c e d u r e , F e t t e r o l f a n d Y o s t ( 1 7 ) h a v e made
c a l c u l a t i o n s o f t h e q u a l i t a t i v e s p e c i f i c i t y o f s e v e r a l mass
s p e c t r o m e t r i c t e c h n i q u e s a n d w e i g h e d them a g a i n s t t h e r e q u i r e m e n t s
o f some a n a l y t i c a l t a s k s . W h i l e t h e i r t r e a t m e n t makes s e v e r a l
a s s u m p t i o n s w h i c h may n o t b e b o r n o u t i n p r a c t i c a l a n a l y s i s , i t
i s u s e f u l f o r gauging a p a r t i c u l a r methodology t o an a n a l y t i c a l t a s k .
F e t t e r o l f a n d Y o s t ( 1 7 ) c a l c u l a t e d t h a t c a p i l l a r y GC/MS a t 1 amu
r e s o l u t i o n u s i n g f u l l mass s c a n s o v e r a mass r a n g e o f 1000 amu
g a v e s u f f i c i e n t i n f o r m i n g power b y i t s e l f t o i d e n t i f y w i t h
c e r t a i n t y a n y c h e m i c a l compound i n a c h e m i c a l m i x t u r e .
In o u r m e t h o d o l o g y , t h e mass s c a n s w e r e , b y n e c e s s i t y , l i m i t e d
t o a t most f i v e i o n s i n o r d e r t o meet t h e m i l e s t o n e f o r
s e n s i t i v i t y . T h i s g r e a t l y r e d u c e s t h e i n f o r m i n g power
o f t h e GC/MS p o r t i o n o f t h e a n a l y t i c a l p r o c e d u r e ; however, t h i s i s
more t h a n c o m p e n s a t e d f o r b y t h e s a m p l e p r e p a r a t i o n s t e p s i n t h e

p r o c e d u r e w h i c h i n c l u d e e x t r a c t i o n , l i q u i d c h r o m a t o g r a p h y and
d e r i v a t i z a t i o n . T h i s i s not wasted i n f o r m a t i o n , but i s n e c e s s a r y t o
i n s u r e t h e q u a l i t a t i v e r e l i a b i l i t y and p r e c i s i o n o f t h e a n a l y s i s
thus f u l f i l l i n g t h e program r e q u i r e m e n t s .
J o h n s o n and Y o s t ( 1 8 ) have d e m o n s t r a t e d t h e r e l a t i o n s h i p between t h e
i n t e n s i t y o f t h e i n s t r u m e n t a l r e s p o n s e t o an a n a l y t e and t h e
s i g n a l - t o - n o i s e r a t i o as a f u n c t i o n o f t h e number o f s t e p s i n an
a n a l y s i s . With each a d d i t i o n a l step i n t h e procedure, t h e magnitude
o f i n s t r u m e n t a l r e s p o n s e d e c r e a s e s due t o u n a v o i d a b l e l o s s o f t h e
a n a l y t e w i t h each m a n i p u l a t i o n as t h e s i g n a l - t o - n o i s e r a t i o
i n c r e a s e s . T h e i n c r e a s e i n t h e s i g n a l - t o - n o i s e r a t i o i s due
p r i m a r i l y t o t h e e l i m i n a t i o n o f i n t e r f e r i n g compounds w h i c h
c o n t r i b u t e t o t h e " c h e m i c a l n o i s e " . T h e o b v i o u s c o n c l u s i o n t o draw
i s t h a t when t r y i n g t o a c h i e v e v e r y l o w l i m i t s o f d e t e c t i o n t h e
a n a l y t i c a l m e t h o d s s h o u l d i n v o l v e a s many s t e p s a s p o s s i b l e b e f o r e
t h e a b s o l u t e s i g n a l l e v e l i s r e d u c e d t o an u n d e t e c t a b l e l e v e l This
approach ignores t h e e r r o r
a sample w h i c h c o u l d p o t e n t i a l l
a n a l y s i s . F i g u r e 3 shows, t h a t w h i l e i n s t r u m e n t a l r e s p o n s e s a r e
g e n e r a l l y l o w e r e d by a d d i t i o n a l s t e p s , t h e o v e r a l l s e n s i t i v i t y
improved as t h e r e l i a b i l i t y o f t h e a n a l y s i s g e n e r a l l y d e t e r i o r a t e s .
T h u s , w h i l e a n a l y t i c a l m e t h o d o l o g i e s d e a l i n g w i t h many s t e p s may
push d e t e c t i o n l i m i t s t o h e r e t o f o r e u n a c h i e v a b l e l e v e l s , t h e y do s o
a t t h e r i s k o f p r o d u c i n g u n r e l i a b l e d a t a . T h e e f f e c t o f human e r r o r
on t h e a n a l y s i s i s p a r t i c u l a r l y d i f f i c u l t t o h a n d l e b e c a u s e s u c h
e r r o r s a r e n o t a m e n a b l e t o m a t h e m a t i c a l m o d e l i n g , h o w e v e r human
e r r o r i s t h e most l i k e l y c o n t r i b u t o r t o t h e r e p o r t i n g o f f a l s e
p o s i t i v e o r f a l s e n e g a t i v e v a l u e s , w h i c h i s t h e most s e r i o u s e r r o r
made i n an a n a l y s i s .
RISK ASSESSMENT
One r e s p o n s i b l e means o f a d d r e s s i n g t h e e f f e c t o f human e r r o r on an
a n a l y t i c a l m e t h o d o l o g y i s t o p e r f o r m a r i s k a s s e s s m e n t on t h e
a n a l y t i c a l p r o c e d u r e . A r i s k / c o s t / b e n e f i t a n a l y s i s i s b a s e d on a
h o s t o f f a c t o r s , i n c l u d i n g r e a s o n i n g , g u e s s work and p a s t
p e r f o r m a n c e . T h e p u r p o s e o f a r i s k a s s e s s m e n t i s t o make s u r e t h a t
c a l a m i t i e s happen f i r s t on p a p e r , n o t i n r e a l i t y . A l t h o u g h a l m o s t
a l l a n a l y t i c a l c h e m i s t s p e r f o r m an i n f o r m a l r i s k a s s e s s m e n t o f some
f o r m o r a n o t h e r , i n a c a s e where t h e a n a l y t i c a l p r o c e s s must be u s e d
t o make d e c i s i o n s o f g r a v i t y , i t i s a p p r o p r i a t e t o b r i n g as much
a n a l y t i c a l f o r m a l i t y as p o s s i b l e t o bear. There i s a w e l l developed
s c i e n c e devoted t o r i s k assessment; i t i s e x t e n s i v e l y used i n t h e
development o f s o p h i s t i c a t e d hardware items o r systems o f l i n k e d
hardware items. Examples i n c l u d e t h e r i s k assessment s t u d i e s which
a r e p e r f o r m e d d u r i n g t h e d e s i g n p h a s e o f a n u c l e a r power p l a n t o r
an a i r c r a f t t o d e t e r m i n e t h e p o s s i b l e modes o f f a i l u r e . Some
o r g a n i z a t i o n s r e q u i r e a r i s k assessment study f o r any development
i t e m r e s u l t i n g f r o m any m a j o r p r o g r a m f o r r e s e a r c h , a e v e l o p m e n t and
a c q u i s i t i o n o f m a t e r i e l items o r systems (19). Although e n g i n e e r s
must p e r f o r m r i s k a s s e s s m e n t s , a n a l y t i c a l c h e m i s t s a r e n o t r e q u i r e d
t o p e r f o r m n o r a r e they knowledgeable o f such r e q u i r e m e n t s , even
t h o u g h a f a i l u r e o f an a n a l y t i c a l m e t h o d o l o g y c a n h a v e i m p l i c a t i o n s
and l o n g r a n g e e f f e c t s a t l e a s t a s s e r i o u s as a s i m i l a r f a i l u r e o f a
hardware item.
To a s s o c i a t e a r i s k w i t h a c o m p l e x a n a l y s i s i t i s n e c e s s a r y t o b r e a k

TABLE I
MASS SPECTROMETRY
DETECTION CRITERIA
LIMIT OF DETECTION: SINGLE ION MONITORING (SIM) PEAK

RETENTION TIME 5 SECONDS
SIGNAL/NOISE RATIO > 5
LIMIT OF VERIFICATION: RETENTION TIME 5 SECONDS.

MINIMUM OF 3 IONS (PREFERABLY
5-7) WITH S/N > 2
LIMITS OF QUANTITATION (RANGE): ESTABLISHED BY HUBAUX-VOS*

UPPER AND LOWER LIMITS BOTH
DEFINED
* Procedure described in Anal Chem 2 (1970) 849-885
I
0 1 2 3 4 5 6 7 8 9 10
NUMBER OF ST EPS
F i g u r e 3. Q u a l i t y o f Data v s . Number o f Steps i n Procedure

down t h e p r o c e d u r e i n t o i n d i v i d u a l a c t i o n s , o r p o t e n t i a l f a i l u r e
causing events. Risk i s u s u a l l y d e f i n e d as:
Risk = P r o b a b i l i t y Severity (20)
The m a g n i t u d e o f r i s k f r o m some e v e n t d e p e n d s on t h e p r o d u c t o f how
o f t e n t h e a n a l y s t t h i n k s an e v e n t w i l l o c c u r and how s e r i o u s l y t h e
e v e n t i m p a c t s on t h e o v e r a l l p r o c e s s . T h e r e f o r e , i t i s i n c u m b e n t on
t h e s c i e n t i s t t o d e v e l o p a q u a n t i t a t i v e s e n s e o f where t h e r i s k s i n
an a n a l y s i s e x i s t , and how s e r i o u s t h e y a r e . The b e s t s y s t e m s
a n a l y s t c a n n o t p e r f o r m t h i s f u n c t i o n ; o n l y t h e p e r s o n who t h e i s
most k n o w l e d g e a b l e a b o u t t h e a n a l y t i c a l p r o c e d u r e can f u n c t i o n as
t h e r i s k a s s e s s o r . T h i s p e r s o n i s n o r m a l l y t h e r e s e a r c h c h e m i s t who
d e v e l o p e d t h e m e t h o d o l o g y and n o t t h e a n a l y s t who may r u n t h e
p r o c e d u r e r o u t i n e l y . He o r she i s most f a m i l i a r w i t h t h e e m e r g i n g
m e t h o d o l o g y and has a b a s i s ( w h e t h e r i t be h i s t o r i c a l , i n t u i t i v e o r
reasoned) to a s s i g n a f a c t o r o f r i s k to the i n d i v i d u a l components of
the a n a l y s i s . T y p i c a l mechanism
i n c l u d e e i t h e r t h e use o
f a i l u r e s and a s s o c i a t e d m i n o r f a i l u r e s w h i c h m i g h t c a u s e them, o r a
" F a i l u r e Modes and E f f e c t s A n a l y s i s M o d e l " (21) w h i c h u s e s l i s t s o f
t h e ways a s y s t e m can f a i l and t h e r e s u l t s o f e a c h f a i l u r e . F o r
t h i s s t u d y , t h e " F a i l u r e Modes and E f f e c t s A n a l y s i s M o d e l " was
chosen.
T h i s r i s k a s s e s s m e n t was c o n d u c t e d on t h e a n a l y t i c a l scheme o u t l i n e d
i n F i g u r e 2. Each s t e p i n t h e method was e x a m i n e d t o i d e n t i f y
p o t e n t i a l e r r o r s t h a t c o u l d o c c u r i f p e r f o r m e d by a " c o m p e t e n t
a n a l y s t " . ( I t s h o u l d be n o t e d t h a t i t i s n o t u s e f u l o r i n f o r m a t i v e
t o do t h i s t y p e o f r i s k a s s e s s m e n t i f one assumes t h e a n a l y t i c a l
p r o c e d u r e i s g o i n g t o be c a r r i e d o u t by an i n c o m p e t e n t o r o v e r l y
n e g l i g e n t i n d i v i d u a l . The f i r s t s t e p i n any q u a l i t y a s s u r a n c e
p r o g r a m i n an a n a l y t i c a l l a b o r a t o r y s h o u l d be t o i d e n t i f y s u c h
i n d i v i d u a l s and e x c l u d e them f r o m p a r t i c i p a t i o n i n any c r i t i c a l
p r o g r a m s . ) W h e r e v e r a p o s s i b l e e r r o r was i d e n t i f i e d , t h e p r o b a b l e
c o n s e q u e n c e o f t h a t e r r o r was a l s o i d e n t i f i e d . Each c o n s e q u e n c e was
r a t e d and p l a c e d i n t o one o f t h r e e c a t e g o r i e s , a g a i n u s i n g t h e
management g u i d e l i n e s as t h e b a s i s f o r t h e s e v e r i t y o f t h e r a t i n g .
The most s e v e r e r a t i n g was c r i t i c a l , w h i c h i n c l u d e d any e r r o r t h a t
could r e s u l t in a t o t a l f a i l u r e of the a n a l y s i s or the r e p o r t i n g of
a f a l s e p o s i t i v e v a l u e . The s e c o n d r a t i n g was s u b c r i t i c a l , w h i c h
was any e r r o r w h i c h m i g h t r e s u l t i n t h e r e p o r t i n g o f f a l s e n e g a t i v e
r e s u l t s . The t h i r d r a t i n g was s e r i o u s , w h i c h was any e r r o r t h a t
might r e s u l t i n poor p r e c i s i o n , poor accuracy or the i n t r o d u c t i o n
o f b i a s i n t o t h e r e s u l t s . In t h e q u a l i t y a s s u r a n c e p l a n , a l l
c r i t i c a l and s u b c r i t i c a l e r r o r s were t r e a t e d as u n a c c e p t a b l e .
S e r i o u s e r r o r s were a c c e p t a b l e i f t h e y o c c u r r e d w i t h i n t h e l i m i t s o f
t h e a d o p t e d q u a l i t y c o n t r o l c r i t e r i a . T a b l e II g i v e s a l i s t i n g o f
the r i s k assessment f o r the a n a l y t i c a l methodology.
QUALITY CONTROL
F o l l o w i n g t h e r i s k a s s e s s m e n t s t u d y , a q u a l i t y a s s u r a n c e p l a n was
d r a f t e d and q u a l i t y c o n t r o l p r o c e d u r e s were i m p l e m e n t e d t o e l i m i n a t e
o r m i n i m i z e t h e p r o b a b i l i t y o f t h e e r r o r s o c c u r r i n g . Each t y p e o f
e r r o r was a d d r e s s e d and w h e n e v e r p o s s i b l e t h e p r o c e d u r e was m o d i f i e d
t o e l i m i n a t e t h e p o s s i b i l i t y o f t h a t e r r o r o c c u r r i n g as i s shown i n
T a b l e I I I . P r o c e d u r a l q u a l i t y c o n t r o l was t h e most e f f e c t i v e t y p e

TABLE II
IDENTIFICATION O F P O S S I B L E E R R O R S
AND P R O B A B L E C O N S E Q U E N C E S
ERROR TYPE PROBABLE CONSEQUENCES RATING
1) STORAGE/TAMPERING TOTAL FAILURE OF ANALYSIS CRITICAL

2) TRANSCRIPTION E R R O R IN S O M E S A M P L E S CRITICAL
3) IMPRECISE S A M P L E HANDLING POOR PRECISION SERIOUS
4) CONTAMINATED REAGENTS F A L S E POSITIVES O R CRITICAL
FALSE NEGATIVES
5) C R O S S CONTAMINATION F A L S E POSITIVES CRITICAL
BETWEEN SAMPLES
6) * * C R O S S C O N T A M I N A T I O N F R O M F A L S E POSITIVES CRITICAL
STANDARD TO S A M P L E S
7) S A M P L E DEGRADATION BIAS (POSITIVE OR NEGATIVE) SERIOUS
(IMPROPER S T O R A G E A N D HANDLING) POSSIBLE F A L S E NEGATIVE SUBCRITICAL
8) DEVIATION F R O M W R I T T E N TOTAL FAILURE OF ANALYSIS CRITICAL
PROCEDURE S Y S T E M A T I C BIAS (POS OR SERIOUS
NEG)

9) MATRIX I N T E R F E R E N C E F A L S E POSITIVES O R CRITICAL
FALSE NEGATIVES SUBCRITICAL

REUTTER ET AL. Risk Assessment as a Tool
TABLE m
IDENTIFICATION O F SPECIFIC QUALITY C O N T R O L
TO BE USED TO ADDRESS EACH TYPE OF PROBABLE
ERROR
ERROR TYPE QUALITY ASSURANCE IMPOSED
1) STORAGE OR TAMPERING STRICT ENFORCEMENT OF

CHAIN OF CUSTODY
2) TRANSCRIPTION MAKE UP ALL LABELS FOR

CONTAINERS IN ADVANCE
CHECK ALL LABELS THREE
TIMES
HAVE INDEPENDENT CHECK OF
LABEL ACCURACY
(SUPERVISORY)
3) IMPRECISE SAMPLE HANDLING USE O F INTERNAL STANDARD
4) CONTAMINATION OF REAGENTS FREQUENT USE O F Q C SAMPLES

AT 0, 20 AND 100 ppb LEVELS
USE OF DISPOSABLE GLASS-
WARE WHENEVER POSSIBLE
USE NEW BATCH OF REAGENT
WITH EACH BATCH OF SAMPLES
5) CROSS CONTAMINATION BETWEEN ALWAYS RUN Q C SAMPLES

SAMPLES BETWEEN UNKNOWNS
NEVER USE SAME CONTAINER
OF REAGENT/SOLVENT MORE
THAN O N C E DURING THE
PROCEDURE
SEQUENCE ALL OPERATIONS
USE DISPOSABLE GLASSWARE
WHENEVER POSSIBLE
6) CROSS CONTAMINATION FROM USE SEPARATE ROOM FOR

STANDARD TO SAMPLE HANDLING STANDARDS WHERE
SAMPLES ARE NEVER ALLOWED
USE STANDARD COCKTAIL OF
TRICHOTHECENES AT KNOWN
RELATIVE CONCENTRATION FOR
ALL QUALITY CONTROL AND
STANDARD SAMPLES.
NEVER USE LIQUID HANDLING
DEVICE FOR MORE THAN ONE
OPERATION IN A BATCH RUN
SEQUENCE ALL OPERATIONS
7) SAMPLE DEGRADATION DETERMINE STABILITY OF

(IMPROPER STORAGE AND HANDLING) SAMPLES DURING METHOD
DEVELOPMENT
INSERT POSITIVE Q C SAMPLES
UPON RECEIPT OF SAMPLES
ADHERE STRICTLY TO SAMPLE
STORAGE PROCEDURES
8) DEVIATION FROM WRITTEN PROCEDURE - C L O S E SUPERVISORY CON-

TROLS
INSERTION OF FREQUENT Q C
SAMPLES WHICH ARE "BLINDS"
TO ANALYST
KEEP RESEARCH CHEMISTS
AWAY FROM ROUTINE ANALYSIS
9) MATRIX INTERFERENCES DEVELOP RUGGED ANALYTICAL

METHODOLOGY
DEVELOP AND ENFORCE STRICT
RULES FOR R E J / A C C E P T . OF
SELECTED ION MONITORING
MASS SPECTROMETRY DATA

to implement i n t h i s study. Q u a l i t y c o n t r o l samples were i n c l u d e d

in the sample stream t o t e s t f o r the o c c u r r e n c e o f c r i t i c a l and
s u b c r i t i c a l e r r o r s and t o m e a s u r e t h e i r m a g n i t u d e .
The Q u a l i t y C o n t r o l C o o r d i n a t o r (QCC) r a n d o m l y d i v i d e d t h e s a m p l e s
i n t o g r o u p s o f f o u r unknown s a m p l e s . One o f t h e f o u r s a m p l e s was
c h o s e n r a n d o m l y f o r r e p l i c a t i o n w i t h i n t h e same g r o u p . Q u a l i t y
c o n t r o l s a m p l e s o f t h e t r i c h o t h e c e n e m i x t u r e were added t o t h e g r o u p
a t t h e 0, 20 and 100 ppb l e v e l . The QCC a r r a n g e d s a m p l e s s u c h t h a t
q u a l i t y c o n t r o l s a m p l e s b r a c k e t e d e a c h unknown s a m p l e , t h e r e b y
e l i m i n a t i n g t h e p o s s i b i l i t y o f c r o s s c o n t a m i n a t i o n between unknowns
i f t h e p r o c e d u r a l q u a l i t y c o n t r o l were f o l l o w e d . The s a m p l e s were
a s s i g n e d new numbers and i n t e r n a l s t a n d a r d s were a d d e d . The a n a l y s t
was g i v e n t h e c o n c e n t r a t i o n o f one o f t h e n o n - z e r o q u a l i t y c o n t r o l
s a m p l e s i n e a c h g r o u p s o t h e q u a n t i t a t i v e a c c u r a c y c o u l d be
monitored v s the c a l i b r a t i o n data in r e a l time.
PERFORMANCE OF THE ANALYTICA

F i g u r e 4 shows a s i n g l e i o n c h r o m a t o g r a m o f t h e b a s e p e a k f r o m t h e
mass s p e c t r a o f t h e n e p t a f l u o r o b u t r y l e s t e r o f s c i r p e n t r i o l
r e s u l t i n g from the e x t r a c t o f a 5 p a r t - p e r - b i l l i o n standard i n
b l o o d . F i g u r e 5 shows t y p i c a l d a t a g e n e r a t e d f o r a n unknown s a m p l e
w h i c h was v e r i f i e d t o have T-2 t o x i n . T h i s c o n f i r m a t i o n r u n was
determined t o be p o s i t i v e by comparison with the responses o f
s t a n d a r d s a t t h e same l e v e l . F i g u r e 6 i s a t y p i c a l r e g r e s s i o n
curve o f c o n c e n t r a t i o n v s the peak h e i g h t r a t i o o f the base peak t o
t h e i n t e r n a l s t a n d a r d , D e o x y v e r r u c a r o l DOV ( 1 , 2 2 - 2 4 ) f o r t h e
e x t r a c t a b l e t r i c h o t h e c e n e s . T a b l e IV l i s t s t h e e x p e r i m e n t a l l y
d e t e r m i n e d l i m i t s o f d e t e c t i o n , v e r i f i c a t i o n and q u a n t i f i c a t i o n f o r
t h i s a n a l y t i c a l t e c h n i q u e when t h e above c r i t e r i a were a p p l i e d .
A n a l y t i c a l c h e m i s t s o f t e n f a c e an a p p a r e n t d i l e m m a i n how t o r e p o r t
a v a l u e which f a l l s above the c r i t e r i a f o r d e t e c t i o n but below the
l e v e l r e q u i r e d f o r v e r i f i c a t i o n , e s p e c i a l l y when t h e p e n a l i t y f o r
r e p o r t i n g f a l s e p o s i t i v e s i s h i g h . One r e s p o n s i b l e and a c c u r a t e
a n s w e r i s t h a t t h e a n a l y t e was i n d i c a t e d a b o v e t h e minimum
d e t e c t a b l e c o n c e n t r a t i o n but c o u l d n o t be v e r i f i e d .
COLLABORATIVE STUDY
The C e n t e r f o r D i s e a s e C o n t r o l (CDC), A t l a n t a , GA was r e q u e s t e d t o
p r e p a r e a s e t o f p r o p e r l y s t a b i l i z e d and d e a c t i v a t e d human b l o o d
s a m p l e s c o n t a i n i n g f o u r s i m p l e t r i c h o t h e c e n e m y c o t o x i n s w i t h two
t o x i n s e a c h f r o m t h e t y p e A and g r o u p s . A s t a t i s t i c i a n f r o m
CRDEC, i n c o l l a b o r a t i o n w i t h CDC, e s t a b l i s h e d a s a m p l e n u m b e r i n g
s y s t e m and a n e x p e r i m e n t a l d e s i g n f o r p r e p a r a t i o n o f t h e s a m p l e s e t ,
which i n c l u d e d at l e a s t f o u r c o n c e n t r a t i o n s of each t o x i n . T h i s
e x p e r i m e n t a l d e s i g n had c h e c k s f o r d i l u t i o n e r r o r and measurement
e r r o r i n m a k i n g up t h e s a m p l e s and p r o v i d e d a s t a t i s t i c a l l y
s i g n i f i c a n t number o f r e p l i c a t e and b l a n k s a m p l e s t o d e t e r m i n e
a c c u r a c y , p r e c i s i o n and t h e e x p e c t e d r a t e f o r r e p o r t i n g f a l s e
p o s i t i v e v a l u e s . The s a m p l e s were s h i p p e d t o CRDEC u n d e r
r e f r i g e r a t i o n . A f t e r CRDEC c o m p l e t e d t h e a n a l y s i s o f t h e s a m p l e s
and r e p o r t e d t h e r e s u l t s t o an i n d e p e n d e n t o r g a n i z a t i o n , CDC
r e l e a s e d t h e s a m p l e key. The r e s u l t s were e v a l u a t e d b y t h e
B a l l i s t i c s Research Laboratory, Aberdeen Proving Ground. The
r e s u l t s a r e s u m m a r i z e d i n T a b l e V.


i i 1 r TfinF F i l * >BS89 580.3 amu Fil* >BS89 581.3 amu.
RDC Ml
14.45 15.13
5000
Il UNKNOWN
4OO0H
i 200
3O0
SOGH 14.46
20
2
X 1000 - Il
0
14.5 15. 15.5 14.5 15.0 15.5 14.5 15.0 H
5
[Fil.? : BS89 622.3 arr.u. S " File > B S 8 9 6 4 2 . 3 a m u . S R M P File >BS89 662.3 n.u .
RDC RDC i
14 . 4 5 15.13 14 . 4 5
2000- 40OOi
1600
1 600-
3O00
120O- 1200-1
15.12 14.46 15^10
2000
800- 800
1000

4 00- 400
0 - f r- 0
f r r
15.0 15.5 14.5 15.0 15.5
14.5 14.5 15.0 15.

F i g u r e 5. M u l t i p l e I o n M o n i t o r i n g o f Unknown
REUTTER ET AL. Risk Assessment as a Tool
CONCENTRATION ppb
VERRUCAROL, 17 AUG 1985, PEAK HT RATIO TO I.S.; DOV= I.S.
F i g u r e 6. C a l i b r a t i o n Curve f o r V e r r u c a r o l
TABLE IV
EXPERIMENTALLY DETERMINED LIMITS
OF ANALYSIS FOR TRICHOTHECENES
IN HUMAN BLOOD
DETECTION VERIFICATION QUANTIFICATION

TOXIN LIMIT* LIMIT RANGE
ppb ppb ppb
T-2 0.5 3 1-200

HT-2 0.5 3 1-50
DAS 0.2 4 1-50
VER .02 0.5 1-20
DON .04 1 1-100
FUS-X 0.1 2 1-100
SCIR 0.1 2 1-100
T-2 TETRAOL .01 0.1 .05-20**
* EXTRAPOLATED FROM LOWEST S/N VALUE.

"DETERMINED AFTER CDC COLLABORATIVE STYDY.

TABLE V
MEANS AND STANDARD DEVIATIONS OF THE
MEASURED DATA IN ORIGINAL UNITS
COLLABORATIVE INTERNAL QUALITY
TOXIN TARGET
200 218.75 9.78 100 102.4 28.4

40 27.25 8.54 20 25.6 10.5
T-2 10 9.88 2.63
2 2.38 .50
0 bdl 0 bdl 0
150
30 NOT
NIV 10 ANALYZED N/A

0
150 177.25 13.05 100 81.6 7.7
30 28.75 2.87
VER 10 15.00 0.00 20 18.6 2.4
2 7.38 0.13
0 bdl 0 0 bdl 0
200 113.25 24.41 100 82.8 14.4

40 21.25 28.87
DAS 10 4.25 0.50 20 18.6 4.2
2 4.25 3.77
0 bdl 0 0 bdl 0
bdl = B E L O W DETECTION LIMIT

= 4
S.D. = Standard Deviation

The r e s u l t s f r o m t h e q u a l i t y c o n t r o l s a m p l e s w h i c h w e r e a n a l y z e d as
p a r t o f t h e CDC c o l l a b o r a t i v e s t u d y a r e shown i n T a b l e s VI and V I I .
F i g u r e s 7 and 8 a r e e x a m p l e s o f c o n t r o l c h a r t s w h i c h show t h e
p r e c i s i o n o f t h e a n a l y s i s t o be a p p r o x i m a t e l y 3 0 % r e l a t i v e s t a n d a r d
d e v i a t i o n f o r any c o n c e n t r a t i o n o t h e r t h a n 0 r e g a r d l e s s o f t h e
identity of the trichothecene. (On a r e s p o n s e v s . c o n c e n t r a t i o n
c h a r t , t h e one s t a n d a r d d e v i a t i o n e r r o r b a r s w o u l d be s e e n a s
d i v e r g i n g f r o m t h e l i n e r e p r e s e n t i n g t h e mean a s t h e c o n c e n t r a t i o n
i n c r e a s e s . I f t h e same d a t a were p l o t t e d as l o g r e s p o n s e v s l o g
c o n c e n t r a t i o n , t h e f i r s t standard d e v i a t i o n e r r o r bars would appear
t o p a r a l l e l t h e c a l c u l a t e d mean.) F i g u r e 9 shows t h e r e s u l t s f r o m
p l o t t i n g t h e f i r s t and s e c o n d r e p l i c a t e s o f CDC s a m p l e s c o n t a i n i n g
T-2 t o x i n . T h i s r e p l i c a t i o n s t u d y d e m o n s t r a t e s t h a t r e a s o n a b l e
e s t i m a t e s o f t h e p r e c i s i o n f o r unknown s a m p l e s c a n be d e t e r m i n e d
from control c h a r t s .
CONCLUSION
A n a l y t i c a l m e t h o d o l o g y was d e v e l o p e d f o r a c c u r a t e q u a n t i t a t i v e
a n a l y s i s o f t r i c h o t h e c e n e s a t low p a r t - p e r - b i l l i o n l e v e l s i n blood.
A l t h o u g h t h i s m e t h o d o l o g y was a r d u o u s and l a c k e d t h e r u g g e d n e s s
n o r m a l l y demanded o f an a n a l y t i c a l p r o c e d u r e w h i c h must nave a l o w
f a i l u r e r a t e s i t p r o v e d t o be b o t h q u a l i t a t i v e l y r e l i a b l e and
q u a n t i t a t i v e l y a c c u r a t e when i t was c o m b i n e d w i t h a w e l l p l a n n e d
q u a l i t y a s s u r a n c e program. An i n d i s p e n s a b l e p a r t o f d e v e l o p i n g t h e
q u a l i t y a s s u r a n c e p l a n was a f o r m a l r i s k a s s e s s m e n t w h i c h
s p e c i f i c a l l y t o o k i n t o a c c o u n t t h e p o s s i b i l i t y o f human e r r o r .
T h i s p r o c e d u r e was v a l i d a t e d by c o l l a b o r a t i v e s t u d y w i t h
independent l a b o r a t o r i e s .
T A B L E VI
RESULTS OF QUALITY CONTROL SAMPLES
TARGET CONCENTRATION VS. REPORTED CONCENTRATION
TARGET REPORTED
SAMPLE # CONCENTRATION CONCENTRATION
(PPb) (PPb)
T-2 VER DAS
QC 2 0 ND* ND ND
QC 6 0 ND ND ND
QC 9 0 ND ND ND
QC 12 0 ND ND ND
QC 13 0 ND ND ND
QC 1 20 21 17 19
QC 5 20 21 20 16
QC 7 20 16 16 14
QC 11 20 27 18 19
QC 14 20 43 22 25
QC 3 100 72 91 67
QC 4 100 82 80 67
QC 8 100 95 82 93
QC 10 100 140 85 93
QC 15 100 123 70 94
*ND = NOT DETECTABLE

TABLE V I I
QUALITY CONTROL SAMPLES
Summary Report
TARGET CONC. STANDARD @ 95% RELATIVE

(ppb) (MEAN) DEVIATION CONFIDENCE LEVEL* ERROR
T-2 20 25.6 10.5 12.6-38.6 28.0%
T-2 100 102.4 28.4 67.1-137.7 2.4%
DAS 20 18.6 4.2 13.4-23.8 - 7.0%
IDAS 100 82.8 14.4 64.9-100.7 -17.2%
VER 20 18.6
VER 100 81.6 7.7 72.0-91.2 -18.4%
USING THE FORMULA FOR CONFIDENCE LIMIT FOR = X t s / v

DEGREES OF FREEDOM = 4.
3 SD
MEAN = 25.6
2 SD
3 SD
SD = STANDARD DEVIATION = 10.2 ppb
-J I I I I L
1 2 3 4 5
F i g u r e 7. Q u a l i t y C o n t r o l Chart

3 SD
2 SD
MEAN = 102.4
2 SD
3 SD
SD = STANDAR
J L
1 2 3 4 5
F i g u r e 8. Q u a l i t y C o n t r o l Chart
200
/
/
IDEAL + 30% RSD- /
y IDEAL
if) /
/
>
/
C
CM / IDEAL - 30% RSD
CL
/
/
/
/
40 80 120 160 200

1st ANALYSIS (ppb)
F i g u r e 9. R e p l i c a t i o n C o n t r o l Study

ACKNOWLEDGMENT
The a u t h o r s g r a t e f u l l y a c k n o w l e d g e M r s . D. A . P a t e r n o , Mr. M. A .
Wasserman a n d D r . T. K r i s h n a m u r t h y f o r t h e i r s u p p o r t i n d e v e l o p i n g
t h e a n a l y t i c a l p r o c e d u r e a n d Mr. L . S t u r d i v a n f o r s u p p l y i n g
s t a t i s t i c a l support t o the project.
1. Krishnamurthy, K.; Sarver, E.W. J . Chromato. 1986, 335,
253-264
2. Kurata, H . ; Ueno, Y. Toxigenic Fungi--Their Toxins and
Health Hazard; Elsevier, New York, 1984.
3. Frank, B. Angew. Chem., Int. Ed. Engl., 1984, 23, 493.
4. Fontelo, P. .; Beheler, J.; Bunner, D. L . ; Chu, F. S.,
Appl. Environ. Microbiol., 1983, 45, 640-3.
5. Pare, J . R. Jocelyn
Apsimon, J . W.
6. Stahr, H. M.; Lerdal, D.; Hyde, W.; Pfeifer, R., Appl.

Spectrosc., 1983, 37, 396-400.
7. Visconti, .; Bottalico, ., Chromatographia, 1983, 17,
97-100.
8. Heyndrikx, A; Sookvanichsilp, N.; Van den Heede, M. Arch.
Belg. Med. Soc., Hyg., Med. Trav. Med. Leg., 1984, Suppl.
143-6
9. Tanaka, T . ; Hasegawa, .; Matsuki, Y.; Ishii, K.; Ueno, Y.,

Food Addit. Contam. 1985, 2, 125-37.
10. Smith, R. D.; Udseth, H. R.; Wright, B. W.; J . Chromatogr.
Sci., 1985, 23, 192-9.
11. Gore, J.; Rougereau, .; Person, O., J . Chromatogr., 1984,
291, 404-8.
12. Wilson, C. .; Everard, D. M.; Schoental, R., Toxicol.

Lett., 1982, 10, 35-40.
13. Protection Against Trichothecene Mycotoxins, National
Research Council, Report, 1983, Order No. AD-A137115, pp238.
14. Rodricks, J . V. ; Hesseltine, C. W.; Mehlman, M. A, Ed.,
Mycotoxins in Human and Animal Health, Pathotox Publishers,
Park Forest, IL, 1977.
15. Mirocha, C. J.; Pawlosky, R. .; Chatterjee, K.; Watson, S.;
Hayes, W.; J. Assoc. Off. Anal. Chem., 1983, 66, 1485.
16. Kaiser, H . , Anal. Chim. Acta, 1978, 33B, 551.
17. Fetterolf, D. D.; Yost, R. .; Int. J . Mass. Spec. and Ion.
Proc. 1984, 63, 33-50.

18. Johnson, J . V.; Yost, R. .; Anal. Chem. 1983, 57, 758a.

19. Army Regulation 70-1, 15 March 1984, Headquarters,
Department of Army, Washington, D. C.
20. Crouch, E. A. C.; Wilson, R.; Risk/Benefit Analysis,
Ballinger, Cambridge, MA, 1982, p9.
21. McKean, K. Discover 1986, 7, 38
22. Schuda, P. F . ; Potlock, S. J.; Wannamacher, R. W. Jr., J.
Nat. Prod., 1984, 47, 514.
23. Jarvis, . B.; Stahly, G. B.; Pavanasasivam, G., Med. Chem.
(Academic), 1980, 23, 1054.
24. Jarvis, . B.; Pavanasasivam, G., Appl. Environ. Microbiol.,
1983, 46, 480.

Chapter 13
Evaluating the Impact of Hypothesis

Testing on Radioactivity Measurement
Programs at a Nuclear Power
Facility
R. A. Mellor and C. L. Harrington
Yankee Atomic Electric Company, Rowe, MA 01367
NUREG 4007, "Lowe

Elaboration of a Proposed Position for Radiological
Effluent and Environmental Measurements" proposed the
application of hypothesis testing to nuclear power
plant Radiological Effluent Technical Specifications
(RETS). Although hypothesis testing has been implicit
in the current RETS, the application of the NUREG
proposal will require a more detailed knowledge of the
measurement system uncertainties, establishment of a
basis for assumed boundary conditions for these
uncertainties and a heightened degree of control over
measurement processes. This paper applies the
requirements of the NUREG to three types of
radiometric measurement systems routinely used for
compliance with the RETS. The methods used to meet
the requirements will be detailed.
Over the past several years Radiological Effluent Technical
Specifications (RETS) have been implemented at nuclear power
f a c i l i t i e s i n the United States. These RETS control the
evaluation and release of radioactive material to the o f f - s i t e
environs. Radiological measurements of these releases are
conducted using gamma spectroscopy, l i q u i d s c i n t i l l a t i o n , gross
counting techniques and continuous on-line monitoring.
The Technical Specifications also contain requirements for
Environmental Monitoring Programs (EMP). These programs are
primarily designed to provide the determination of dose,
assessment of trends of r a d i o a c t i v i t y i n the environment and
public reassurance (1). Many of the same r a d i o l o g i c a l measurement
techniques are used for environmental monitoring but with the
intent of detecting r a d i o a c t i v i t y orders of magnitude less than
effluent release source terms.
The concept of hypothesis testing has been applied to
o f f - l i n e r a d i o l o g i c a l measurement techniques used i n the RETS for
both in-plant and environmental monitoring. S t a t i s t i c a l l y v a l i d
m u l t i p l i e r s , the random v a r i a t i o n of the background, and
0097-6156/88/0361 -0244$06.50/0

13. MELLOR & HARRINGTON Impact of Hypothesis Testing 245
measurement protocol (MP) variables have been combined i n standard

radiometric formulations to provide an estimation of the lower
l i m i t of detection (LLD) for the MP. Numerical LLD values have
been assigned as requirements which the MP must achieve f o r
various radionuclides. Generally these requirements are set to
ensure the release of radionuclides below these detection l i m i t s
would have a n e g l i g i b l e e f f e c t on the health and safety of the
public.
NUREG 4007 (2) (hereafter c a l l e d the NUREG) represents a
d e f i n i t i v e treatment of the LLD concept. The basic change from
the current RETS approach recommended i n the NUREG i s the
incorporation of upper bound r e l a t i v e systematic uncertainty terms
in the determination of the LLD. This approach would allow the
LLD determination to include the influence of any p o t e n t i a l
systematic uncertainty i n either the background determination or
MP variables. E x p l i c i
of actual upper bound value
uncertainties. Currie, , provide
considers reasonable values f o r these systematic uncertainty
bounds which represent "routine state-of-the-art" i n radiometric
measurements. However, there are instances within the NUREG where
the substitution of s i t e s p e c i f i c values are recommended and where
more c a r e f u l evaluation of these bounding conditions by the
Nuclear Regulatory Commission (NRC) i s l e f t open to consideration.
This work evaluates the systematic errors possibly present i n
three MPs, compares these uncertainty bounds to those used i n the
NUREG and applies the r e s u l t s of these evaluations i n a comparison
of current and proposed LLD formulations. The measurement
protocols chosen for t h i s evaluation are: t r i t i u m (H-3) analysis
in power plant l i q u i d e f f l u e n t s , low l e v e l 1-131 analysis i n milk,
and gamma spectroscopy of power plant l i q u i d e f f l u e n t s . These
protocols exhibit routinely achieved LLDs which are ten to twenty
percent of the regulatory requirements.
Hypothesis Testing and LLD Formulations
The use of hypothesis testing to define the LLD has been evaluated
previously (1-4). Two states of any measurement system composed
of normally d i s t r i b u t e d random uncertainties are considered: the
n u l l hypothesis state i n which the samples contain no net
r a d i o a c t i v i t y and the d i s t r i b u t i o n about the net count of zero i s
characterized by the Mean ( y ) and the standard deviation
Q
( a ) ; and, the LLD state (Currie's "alternate hypothesis" i n

0
the NUREG) i n which the d i s t r i b u t i o n about the net counts at the

LLD i s characterized by the mean ( y ) and the standard
D
deviation (<*) The ultimate question for any sample data i s

"which state i s most consistent with the data?". In making t h i s
decision, there i s a chance that we w i l l f a l s e l y conclude the data
i s part of the d i s t r i b u t i o n about the LLD or that we w i l l f a l s e l y
conclude the data i s part of the d i s t r i b u t i o n about the net count
of zero. These r i s k s are defined by the p r o b a b i l i t i e s and
respectively. These r i s k p r o b a b i l i t i e s may be chosen at any l e v e l

but have been chosen f o r the purposes of regulatory LLD

requirements at f i v e percent. This leads to the regulatory
d e f i n i t i o n of LLD.
LLD - The smallest concentration of radioactive

material i n a sample that w i l l y i e l d a net count,
above system blank that w i l l be detected with at least
a 95 percent p r o b a b i l i t y with no greater than a 5
percent p r o b a b i l i t y of f a l s e l y concluding that a blank
observation represents a " r e a l " s i g n a l .
Based upon t h i s d e f i n i t i o n the current RETS formulation of LLD can

be derived (2,4) and i s shown below.
4.65 S. (1)
_ . _ _ D
Where:
=
detection e f f i c i e n c y
V =
sample size (mass or volume)
Y =
radiochemical y i e l d
2.22 =
dpm per picocurie
=
r a d i o a c t i v i t y decay constant
for the p a r t i c u l a r radionuclide
t = elapsed time between c o l l e c t i o n
and analysis
Sfc = standard deviation of the background
counting rate or the counting rate
of a blank sample as appropriate
The NUREG formulation of the LLD (shown below) incorporates

systematic uncertainty boundary conditions f o r , V, Y, the
counting time (T) and the background (B). The reader i s directed
to the NUREG f o r the complete rationale behind the manner of
incorporation of these uncertainty bounds.
LLD = f ( 2 l B + 2.71 + 3.29<,)

B q ( 2 )
2.22 (YEVT)
f = an amplification factor providing conservative
bounds for systematic uncertainty i n , , V
or (and = 1 + A^)
i B = r e l a t i v e systematic bound i n the blank

( i s i n counts)
oq = standard deviation of the true net

signal
Assigning f = 1.1, A = 0.05 and equating

B
<J = <*/ y i e l d s the

0
following:

IS. MELLOR & HARRINGTON Impact of Hypothesis Testing 247
1 ( 2 7 1 3 2 9 )
LLD = 0.11*BEA + ' + ' V (3)
2.22 (YEVT)
Where:
= standard deviation of the blank
= 1+1/b where b = r a t i o of the blank and sample

counting times
= Blank Equivalent A c t i v i t y
B/2.22 (YEVT)
NOTES: 1. If B>70 counts the constant of 2.71 i s not required

2 2 2 2 2
2. A A = Ay + A E + v +i T
The r e l a t i v e systemati
t h i s evaluation.
Methods of Evaluation
In preparing to determine the r e l a t i v e systematic uncertainty

bounds associated with an a - p r i o r i LLD, i t was decided to use
methods which, although might not be rigorous from a s t a t i s t i c a l
perspective, would allow the estimation of these upper bounds i n a
r e l a t i v e l y easy manner. Four methods of evaluation have been used
in t h i s study.
Method I i s a "theoretical** approach (5) which uses
systematic boundary conditions for each of the parameters of the
MP coupled with appropriate propagation techniques to arrive at an
o v e r a l l systematic uncertainty. In establishing certain of these
boundaries, the central tendancy of the parameter was estimated,
wherever possible, by determining the mean of a minimum of sixty
observations of the parameter i n question. The systematic
uncertainty bound was established by determining the confidence
i n t e r v a l f o r the 95.45 percent confidence c o e f f i c i e n t of the
d i s t r i b u t i o n (2 s ) and expressing t h i s value as a percentage of
x
the mean. The use of the confidence i n t e r v a l at i n f i n i t e degrees

of freedom i s estimated to induce only a two percent uncertainty
in the determined quantity versus using the student-t s t a t i s t i c .
In Method I I , the r e l a t i v e systematic uncertainty i n other
parameters such as the y i e l d were determined by propagating the
uncertainties quoted from manufacturer l i t e r a t u r e f o r balances and
automatic pipetting equipment with the experience estimates f o r
uncertainties of c a r r i e r content, stable element contributions and
variations i n f i n a l p r e c i p i t a t e weights. Estimation of the
boundary conditions f o r e f f i c i e n c y determinations were, i n some
instances, based upon the maximum deviation from the mean response.
The r e l a t i v e systematic uncertainty i n the background was
estimated (Method I I I ) , where possible, from background
d i s t r i b u t i o n s having a minimum of 120 observations. The
background d i s t r i b u t i o n s were evaluated for normality using the
standard chi-square test. The determination of the r e l a t i v e
systematic uncertainty was made by assuming a p e r f e c t l y normal
Library
155 16th St., N.W.
ACS Symposium Series; Washington, D.CSociety:
American Chemical 20036Washington, DC, 1987.
d i s t r i b u t i o n and calculating the difference between the observed

standard deviation and the theoretical standard deviation. Two
times the absolute value of the f r a c t i o n a l difference divided by
the mean [ 2 | ( S - / | ] was considered to be the
N 0
upper bound f o r the systematic uncertainty i n the d i s t r i b u t i o n .

Method IV i s empirical and involves the evaluation of
intra-laboratory knowns ( 5 ) . The systematic uncertainty may be
estimated by comparing the mean experimental result with the known
value i n accordance with the equation below.
i = e + 6 M
Where:
e = the t o t a l uncertainty (mean-known)
6JJ = two times the standard error of the mean.
As much indpendance fro

maintained during thes recognized, ,
that the laboratory dpendance of these measurements may l i m i t
t h e i r usefulness for evaluation purposes. These techniques were
applied, as appropriate, to each of the three MPs under
consideration. The detailed evaluations of each technique are
presented below.
Tritium (H-3)
The t r i t i u m measurement protocol i s r e l a t i v e l y s i m p l i s t i c and

consists of n e u t r a l i z a t i o n of the sample, single plate
d i s t i l l a t i o n , removal of an appropriate size aliquot (usually
eight mL.) v i a reproducible automatic pipets and the addition of
1 5 mL. of dark adapted s c i n t i l l a t i o n c o c k t a i l under incandescent
l i g h t i n g conditions. After shaking to ensure a uniform g e l , the
sample i s allowed to s e t t l e and dark adapt f o r up to twenty but no
more than s i x t y minutes. The l i q u i d s c i n t i l l a t i o n unit e f f i c i e n c y
i s determined d a i l y on a previously prepared standard and
background measurements are determined at least d a i l y . Quench
corrections are not applied to the system due to the lack of an
external standards r a t i o c a p a b i l i t y and an e f f o r t to minimize the
amount of hazardous waste which would be generated i f an internal
standards approach were adopted.
Over a f i v e month period, approximately 6 0 data points were
collected f o r the e f f i c i e n c y and 1 5 0 data points f o r background
determinations. The frequency d i s t r i b u t i o n for the e f f i c i e n c i e s
i s shown i n Figure 1 . The r e l a t i v e standard deviation f o r t h i s
d i s t r i b u t i o n i s 0 . 0 1 1 and the r e l a t i v e systematic uncertainty
bound determined i n accordance with Method I i s 0 . 0 2 2 .
The e f f e c t of the lack of a quench correction was evaluated
by reviewing the same technique accomplished at a d i f f e r e n t
laboratory within the Corporation but with the added protocol step
of e f f i c i e n c y determination v i a the internal standard method. A
t o t a l of 6 6 data points comprise the d i s t r i b u t i o n depicted i n
Figure 2 . This data i s considered normally distributed with
2SQ = 0 . 0 7 1 . These two r e l a t i v e systematic uncertainty boundary
conditions f o r the e f f i c i e n c y have been propagated i n quadrature
to y i e l d a r e l a t i v e systematic uncertainty of 0 . 0 7 4 .

S.
tu
>
w
C*
55*
m
5'
S
>
8
H


1
a
>
S
>
w
>


The e f f e c t of the sample standing (dark adapting) time on the

e f f i c i e n c y of detection of H-3 i n a sample has been evaluated v i a
a series of three experiments. Samples prepared i n the routine
manner were immediately placed i n the well of the l i q u i d
s c i n t i l l a t i o n counter and a one minute analysis i n i t i a t e d . The
sample was then re-analyzed at approximately f i v e minute intervals
and the r a t i o of the observed counts at time t=x to those at time
t=0 were determined and plotted versus the elapsed time x. The
r e s u l t s , i l l u s t r a t e d i n Figure 3., indicate a 5.3 percent average
bias (n=17) a f t e r t h i r t y minutes which remains r e l a t i v e l y
constant. The bias averages (n=13) 3.8 percent a f t e r twenty
minutes. The r e l a t i v e standard error of the mean (RSEM) for the
zero to twenty minute data (0.004) i s equal to the RSEM f o r the
t h i r t y to s i x t y minute data. The ranges of the two means do not
cross indicating the difference i s r e a l . Since samples are
routinely held f o r twent minute p r i o t analysis th
difference between the
systematic uncertainty
been combined i n quadrature with the r e l a t i v e systematic
uncertainty previously determined to obtain the r e l a t i v e
systematic uncertainty bound f o r the e f f i c i e n c y of 0.076.
At the end of 55-60 minutes, the maximum deviation of the
individual t r i a l ratios was two percent. This minor deviation has
been assigned to v a r i a b i l i t y i n sample preparation and has been
added i n quadrature with the automatic pipet l i t e r a t u r e precision
estimate of one percent to a r r i v e at the r e l a t i v e systematic
uncertainty bound for the volume of 0.022. The t o t a l r e l a t i v e
systematic uncertainty bound f o r the efficiency-volume term i n the
denominator of the LLD equation w i l l therefore be 0.079. This was
calculated by adding the r e l a t i v e systematic errors i n quadrature.
The background frequency d i s t r i b u t i o n for the l i q u i d
s c i n t i l l a t i o n unit i s graphically portrayed i n Figure 4. The
d i s t r i b u t i o n was determined to be normal. The value of

2 I s (H-3) ~ (H-3) |/ (Method III) was
B
determined to be 0.004 based upon y = 22.59 cpm and S =

B B
1.46 cpm for a ten-minute analysis. The results of the

evaluations of r e l a t i v e systematic uncertainty bounds have been
tabulated i n Table I.
1-131 Low Level Analysis
Low l e v e l (LLD < 1 pCi/Kg) 1-131 measurements have been conducted

at the Yankee Atomic Environmental Laboratory (YAEL) for over
eight years. Chemical separation, p r e c i p i t a t i o n and mounting of
Cul, and analysis on beta/gamma coincidence systems form the basis
of the technique (6). The stable iodide concentration i s
determined by an ion selective electrode method. The large volume
milk samples are weighed on top-loading balances and stable
element iodide c a r r i e r i s added v i a control checked automatic
pipets. After e q u i l i b r a t i o n of the c a r r i e r , the sample i s
processed through an anion ion exchange column. The stable iodide
present on the column i s subsequently removed; and through a

1.1
1 .09
1 .08
1 .07

1 .06
1 .OS
1 .04-
D
1 .03
1 .02
1 .01
1 I
2 0 40 60

Time (mln) Sine* Sample n Counter
Figure 3 E f f e c t of Sample S e t t l i n g Time on Tritium Count Rate

(CR). T r i a l 1 - ; T r i a l 2 - +, T r i a l 3 -

e*
*
S,
en
r<
5'
>
.
NH
S.
e*
PO

Co
W1

TABLE I
ESTIMATES OF THE UPPER BOUNDS

OF RELATIVE SYSTEMATIC UNCERTAINTIES
UNCERTAINTY MEASUREMENT PROTOCOLS

TYPE H-3 1-131 (LL) GAMMA
*E 0.076 0.075 0.068
*Y
0.079 0.106 0.070
0.004 0.039
NOTE: has been considered n e g l i g i b l e f o r these evaluations.
* Not able to be determined from available gamma spectroscopy

information.
series of oxidation/extraction/reduction steps, the iodine present

i s isolated as an iodide solution and Cul i s p r e c i p i t a t e d by the
addition of an a c i d i f i e d CuCl solution. The Cul i s f i l t e r e d ,
mounted and dried to constant weight. The chemical y i e l d i s
determined based upon the t o t a l weight of stable iodide i n the
sample (stable present i n the milk plus c a r r i e r ) . The p u r i f i e d
sample i s analyzed on a beta/gamma coincidence analysis system
which detects coincident 1-131 beta and gamma emissions from the
sample.
A large data base has been amassed r e l a t i v e to the analysis
of milk samples f o r 1-131. The data are presented here along with
the " t h e o r e t i c a l " estimates of systematic uncertainty i n each of
the parameters of concern i n order to provide the reader with a
review of the type of data required f o r experiment redesign and
optimization. This h i s t o r i c a l data might be used to e s t a b l i s h an
a - p r i o r i LLD estimate of a measurement system.
The units of reporting environmental 1-131 concentrations i n
milk at the YAEL are pCi/Kg and thus, the sample mass i s
determined f o r each sample processed. The h i s t o r i c a l mass data
are portrayed i n Figure 5 and the spread of data shown i s
primarily dependent on the exact amount of milk submitted f o r


Is
h
S
en
en
*.
2.
.

*
Co
Co
3

H

analysis. The evaluation of t h i s data determined a mean

(yy) = 3.598 Kg and a standard deviation (<J ) = 0.165 Kg. w
For any individual measurement, a maximum uncertainty i n the

weight would be 0.5 percent based upon experience.
The chemical recovery of Cul f o r t h i s system i s determined
gravimetrically. Each of the weight terms has an associated
uncertainty which are l i s t e d below. These uncertainties are based
upon the p r e c i s i o n allowed (e.g. i n a c a r r i e r c a l i b r a t i o n ) , the
maximum uncertainty i n measuring the stable iodide concentration
and an estimate of the e f f e c t of v a r i a t i o n of the Cul
p r e c i p i t a t i o n solution pH causing the co-precipitation of CuCl.
[mg. of iodide added (20.0)] = 0.02

[mg. of iodide present (5.0)] =0.10
[mg. of Cul recovered (29.41)] = 0.02
The values i n parenthese

condition for the quantit
average h i s t o r i c a l chemical recovery of 78.42 + 5.38 percent.
Propagation of the l i s t e d uncertainties i n accordance with
established s t a t i s t i c a l methods yields a r e l a t i v e systematic
standard deviation (a /R) = 0.037. The frequency
R
d i s t r i b u t i o n s f o r the chemical recoveries exhibited by two

chemists are portrayed i n Figure 6. The main point to be
remembered from these two d i s t r i b u t i o n s i s that no two individuals
w i l l have the same d i s t r i b u t i o n of r e s u l t s . Establishing an
a - p r i o r i LLD based upon the results of a single individual may
not be applicable to other individuals.
The e f f i c i e n c y of detection for any individual sample i s a
function of the c a l i b r a t i o n i n use at the time of sample
processing. Any c a l i b r a t i o n based upon the preparation of sources
from one stock solution, one c a r r i e r preparation and/or one series
of instrument settings must, by i t s very nature, be considered to
be a new baseline for systematic uncertainty (2). Thus, only by
comparing the differences between c a l i b r a t i o n s can the upper bound
of the systematic uncertainty be determined. Data was available
for the c a l i b r a t i o n of a beta/gamma coincidence unit over a
five-year span of time. The maximum f r a c t i o n a l deviation of any
single e f f i c i e n c y at the average p r e c i p i t a t e weight from the mean
e f f i c i e n c y i s 0.075. This w i l l be used as the upper bound of the
systematic uncertainty i n the e f f i c i e n c y .
Two times the volume and recovery r e l a t i v e systematic
uncertainties were added i n quadrature with the maximum deviation
in the e f f i c i e n c y to provide the boundary condition f o r the
quantity 1 with a result of 0.106. This value i s consistent
A
with the value assumed i n the NUREG.

The background frequency d i s t r i b u t i o n for a beta/gamma system
i s depicted i n Figure 7. This d i s t r i b u t i o n i s composed of 175
800-minute determinations of the system background spanning f i v e
d i f f e r e n t c a l i b r a t i o n s . Any e f f e c t due to the d i f f e r e n t
calibrations has been assumed n e g l i g i b l e . The data are normally
d i s t r i b u t e d . The four highest data points were rejected based
3 6 2
upon Chauvenet's c r i t e r i a [\72 ~ - > 2.98]. The

ES
i i i

>
m
(

"
o
10 )
303*

l-o

ES
- h
z
k\\\\\\ v
2
10 0

UJ

D

k\\\\\\\\\\\ v
[\\\\\\\\\\\\\\^^- !
K\\\\\\\\\\ v
.

IWWWWWW^
T"
10
o 10 CM
A3N3003U

estimation of the r e l a t i v e systematic uncertainty bound i n the

background was determined as:
2 Is B (1-131) - (1-131) |/ * 0.039.

This value was based upon y = 0.0236 cpm and

B
s = 0.0059 cpm for an 800-minute count. The systematic

B
uncertainty boundary conditions f o r the low l e v e l iodine-131

method are provided i n Table I.
Gamma Spectroscopy
This section deals with the process of determining the

photon-emitting components of l i q u i d r a d i o l o g i c a l releases from
pressurized water nuclear power f a c i l i t i e s . The method of
performing l i q u i d effluen
Station (YNPS) involve
followed by immediate analysis for photon emitters i n a 3.5 L
geometry. The system used for analysis i s a "state-of-the-art**,
commercially available, gamma spectroscopy system including
13-15 percent r e l a t i v e (to Nal at Co-60) e f f i c i e n c y high purity
germanium semi-conductor photon detectors housed i n poured
four-inch thick lead shields, high resolution spectroscopy
amplifiers coupled to 500 MHz analog to d i g i t a l converters (ADC)
and a multi-channel analyzer for data acquisition. Spectra are
analyzed v i a vendor software to reduce the observed signals to
pCi/mL concentrations.
Estimates of batch release concentrations are performed
without decay correction since the exact time of the release i s
not known at the time of the measurement, the release i s i n the
near future (usually less than s i x hours from analysis) and
f a i l u r e to incorporate a decay correction maximizes the reported
a c t i v i t y f o r any isotope. Samples are analyzed f o r a length of
time (usually 5,000 seconds) which w i l l ensure the LLD
requirements are achieved f o r each individual sample. This
concept must be incorporated since the YNPS RETS require
n o t i f i c a t i o n of the NRC, on a yearly basis, of a i l samples for
which the LLD has not been achieved. This concept i s tantamount
to establishing the l a s t spectral evaluation of the sample i n
question as the latest a - p r i o r i LLD estimate for the sample under
investigation ( i . e . the analysis incorporates a l l of the
interferences s p e c i f i c to that sample at that time) (2).
The volume analyzed f o r each individual sample i s a t i g h t l y
controlled variable. Although the f o u r - l i t e r m a r i n e l l i beakers (a
water containing volume having a center cavity which w i l l house
the photon detector) used i n the analysis are currently f i l l e d to
a specified height (14.1 cm), the allowed variation does not (by
protocol) exceed 0.2 cm. For chemically pure water specimens,
t h i s translates to a maximum allowable difference of 1.4 percent
by volume. This maximum allowable uncertainty has been used for
the f i n a l estimation of the r e l a t i v e systematic uncertainty i n the
volume term of the a c t i v i t y or LLD equation.
The e f f i c i e n c y of detection of photon emitters i n the four L
m a r i n e l l i beaker geometry has only been determined three times
since the system has been functional. The maximum r e l a t i v e

variation (difference) i n the observed e f f i c i e n c i e s has been

u t i l i z e d as the r e l a t i v e systematic upper uncertainty bound for
the e f f i c i e n c y . A t o t a l of 21 points were involved i n the
evaluation with the maximum f r a c t i o n a l difference for any three
points at the same energy being 0.068.
The combined ( i n quadrature) r e l a t i v e systematic uncertainty
bounds for and V have been determined to be 0.070. This
r e l a t i v e systematic uncertainty i s within the boundary condition
established within the NUREG. I t should be stressed that the
areas evaluated represent only a portion of the a n a l y t i c a l
evaluation performed by the current "state-of-the-art'* software
systems. Peak search and complex spectral f i t t i n g algorithms have
not been addressed d i r e c t l y to date i n t h i s evaluation. An
attempt w i l l be made to address some of these items i n a l a t e r
section through evaluation of samples containing added isotopes of
known quantity.
The estimation o
background i s a much mor
which do not have f u l l knowledge of the algorithms employed or the
exact background count rates achieved for various areas of the
photon spectrum during LLD estimations. I f , however, the
measurement protocols under investigation i n t h i s study are
rigorous with respect to the required LLDs then they should be
able to r e l i a b l y predict the concentration of an analyte present
at the required LLD. This approach has been applied to the gamma
spectroscopy system at the YNPS. A review of the systematic
uncertainty boundary conditions for gamma spectroscopy i s provided
in Table I.
A t y p i c a l l i q u i d release was chosen at random to be evaluated
in d e t a i l with regard to the LLD for two s p e c i f i c isotopes -
Ce-144 and Ru-106. Cerium -144 and Ru-106 were analyzed not to be
present i n the sample at one-tenth the LLD requirements. The
intention of t h i s experimental study was to evaluate the gamma
spectroscopy system response when presented with the presence of
Ce-144 and Ru-106 at the estimated LLD and at the required LLD.
In order to determine the l e v e l at which the isotopes should be
added, an experiment was conducted to determine the average LLD
for both isotopes over a r e l a t i v e l y short, three day, period of
time. The isotopes determined to be present in t h i s release and
t h e i r approximate concentrations (yCi/mL) at time t = 0 are
1-131 (7E-8), Cs-134 (5E-8), Cs-137 (7E-8), Co-60 (2E-8), Kr-85
(7E-6) and Xe-133 (2E-4). One potential problem which was noted
in the NUREG i s interestingly portrayed i n Figure 8. As can be
observed from an evaluation of the LLD versus days from sampling,
the LLD decreases for Ce-144 but remains ( r e l a t i v e l y ) stable over
the time period of the measurement for Ru-106. This reduction i n
observed LLD i s possibly a function of the decreasing 1-131
concentration i n the sample which would, most probably, produce
the largest variation i n the Ce-144 LLD.
After repeated evaluation of the sample, a known amount of
Ce-144 equivalent to the average observed LLD value of Ce-144 was
added to the sample. This amount was approximately equal to 1.2
times the system LLD at the midpoint of subsequent evaluations.
Subsequent to ten determinations (results reported on Table II) an
additional amount of Ce-144 was added to the sample to bring the

W

>
tel


Days From Sampling to Analysis
F i g u r e 8 V a r i a t i o n o f Ce-144 () and Ru-106 (+) Lower L i m i t s

o f D e t e c t i o n w i t h t h e Age o f L i q u i d Waste.
TABLE II
EXPERIMENTAL RESULTS OF KNOWN ADDITIONS TO A ROUTINE

RETS LIQUID MATRIX
Ce-144 Ru-106
At 1.2xSystem At Required At System At Required
LLD LLD LLD LLD
Known Value
(yCi/ml) 1.37E-7 5.00E-7 1.34E-7 5.00E-7
Number of
Measurements 10 25 10 13
Number
Detected* 10 25 9 13
(yCi/ml) 1.28E-7 5.15E-7 1.46E-7 5.06E-7
s
(yCi/ml) 0.24E-7 0.37E-7 0.24E-7 0.70E-7
Number Within 10 25 10 12
X+ - 2s
RSD Ranges f o r
Individual
Measurements 0.15-0.24 0.06-0.07 0.19-0.26 0.10-0.13
S (net counts) 115 454 36 123
* s (counts)
Detected by peak 20 30at routine operational
search algorithm 5.9 17
settings.
NOTE: At the system LLD S > 3.3s indicating a possible bias of

the LLD algorithm.

known l e v e l of Ce-144 to the required LLD concentration

(5E-7 pCi/mL). A t o t a l of 25 more measurements of the
a c t i v i t i e s i n t h i s sample were obtained. The pertinent results
are portrayed i n Table I I . During the measurements of Ce-144,
Ru/Rh-106 was added to the sample. F i r s t the addition was made at
the system quoted LLD and l a t e r (after ten subsequent
measurements) an addition was made to bring the l e v e l up to the
required LLD f o r other isotopes (5E-7 pCi/mL). The point should
be made that the measurement protocol and sample count time
remained the same throughout a l l of the measurements. Table I I
depicts results which are indeed encouraging. The maximum
deviation of the mean from the known was nine percent with a
minimum of ten measurements. The use of this methodology at the
Yankee Atomic E l e c t r i c Company (YAEC) should not be taken as
condoning the use of known value "samples" at the quoted LLD of a
licensee. This approac
basis and could only r e s u l
detected or not detected , analysi sampl g
known concentration of a p a r t i c u l a r isotope at the required LLD
and t y p i c a l interferences should be detected greater than nine out
of ten times i f the requirement to u t i l i z e an i n i t i a l evaluation
of the sample matrix as an a - p r i o r i LLD estimate i s imposed.
Regardless of the regulatory approach, i t has been

demonstrated that analysis of two of the most problematic nuclides
in gamma spectroscopy, at levels well below those deemed harmful
to the public can be accomplished and can provide protection to
the public without recourse to a revised formulation of the LLD
equation.
Potential Impact of the Proposed NUREG
The e f f o r t extended to e s t a b l i s h the appropriateness of the NUREG

assumed (or assigned) r e l a t i v e systematic uncertainty bounds has
been substantial f o r the three techniques under evaluation i n this
study. Several man-months have been expended i n c o l l e c t i n g ,
analyzing and portraying the data f o r the three techniques. These
three techniques represent only 15-20 percent of the techniques
which may require v a l i d a t i o n under the proposed RETS changes. A
d e f i n i t i v e estimate of the time involved would, perhaps, predict
one-two man years of e f f o r t per f a c i l i t y . This estimate may
indeed be placed higher i f the number of on-line monitors i s
substantial i n any single u t i l i t y . However, t h i s e f f o r t might be
considered worthwhile i f the health and safety of the public were
indeed improved as a r e s u l t of the e f f o r t s undertaken by each of
the u t i l i t i e s (estimated from above as approximately two hundred
man-years of e f f o r t ) .
In order to define the p o t e n t i a l improvement to the health
and safety of the public, the actual LLDs calculated under the
current RETS, proposed RETS with the assumed r e l a t i v e systematic
uncertainty bounds and the proposed RETS with calculated r e l a t i v e
systematic uncertainty bounds have been determined f o r two of the
robust techniques under consideration: t r i t i u m analysis and low
l e v e l 1-131 analysis. The t h i r d technique, gamma spectroscopy,
has been shown to be robust with respect to the LLD requirements

via the method of "known addition". As can be seen from Table III
even under the worst set of circumstances, the calculated LLD i s
increased only to within a factor of two of the required LLD. It
must be remembered that the calculations are based upon the
boundary conditions and may not be t r u l y i n d i c a t i v e of the routine
conditions of the measurement process. If a s i g n i f i c a n t change i n
boundary conditions i s noted, a s t r i c t application of the NUREG
concepts might require re-evaluation of the LLD parameters.
TABLE III
COMPARISONS OF CALCULATED LLDs

FOR VARIOUS FORMULATIONS OF THE LLD
(CAS )
RETS (CASE 2) LLDs
H-3 1.79E-6 2.60E-6 1E-5

yCi/ml 2.03E-6
1-131 5.50E-2 7.30E-2 1

pCi/Kg 7.30E-2
Case 1: Based upon the NUREG-4007 r e l a t i v e systematic

uncertainty boundary conditions.
Case 2: Based upon the estimated r e l a t i v e systematic

uncertainty boundary conditions from this study.
The t r i t i u m analysis case was further evaluated i n accordance

with Method IV by preparing a tap water sample (from well water)
with a t r i t i u m standard so the r e s u l t i n g concentration was equal
to the required LLD (1E-5 yCi/mL). The sample has been analyzed
a t o t a l of twenty-six times under the standard protocol with the
following r e s u l t s : y = 1.08+0.06E-5 yCi/mL. Each sample
s
analysis would have predicted (based upon the experimentally

determined NUREG LLD) the presence of t r i t i u m at the calculated
c r i t i c a l l e v e l (1.3 E-6 yCi/mL). Although more analyses may be
required to demonstrate with absolute assurance that the t r i t i u m
methodology i s capable of detecting (and quantitating) t r i t i u m at
the LLD requirement, the implication from the data i s clear: the
public i s well protected by these protocols i f the achieved LLDs
t r u l y r e f l e c t the concentrations which have been established to
protect the public (e.g. H-3 LLD = 1E-5 yCi/mL).

Acknowledgments
The authors g r a t e f u l l y acknowledge the help of M. W. T h i s e l l and

Mark J . Smith of the YNPS for t h e i r help i n c o l l e c t i n g and
portraying data and the s t a f f of the YAEL f o r t h e i r support of
t h i s project.
Literature Cited
1. Upgrading Environmental Radiation Data, HPSR-1 (1980),
Watson, J. E . , Chairman, August 1980.
2. Currie, L. ., NUREG/CR-4007, Lower Limit of Detection:
Definition and Elaboration of a Proposed Position for
Radiological Effluent and Environmental Measurements,
September 1984.
3. Altshuler, B. and Pasternak B. Health Physics 1963 9
293-298.
4. Currie, L. ., Anal , , ,
5. Currie, L. A. and DeVoe, J. R., Systematic Error in Chemical
Analysis, American Chemical Society, Washington, DC, 1977,
114-139.
6. McCurdy, D. E.; Mellor, R. .; Lambdin, R. W.; and McLain,
M. E . , Health Physics, 1980, 38, 203-213.

Chapter 14
Radioactivity Analyses and Detection

Limit Problems
of Environmental Surveillance
at a Gas-Cooled Reactor
James E. Johnson and Janet A. Johnson
Department of Radiology and Radiation Biology, Colorado State

University, Fort Collins, CO 80523
The Lower Limit of Detection (LLD) values required by

the USNRC for nuclear power facilities are often dif-
ficult to attain even using state of the art detection
systems, e.g. the required LLD for I-131 in air is 70
3
fCi/m . For a gas-cooled reactor where I-131 has never
been observed in effluents, occasional false positive
values occur due to: counting statistics using high
resolution Ge(Li) detectors, contamination from
nuclear medicine releases and spectrum analysis sys-
tematic error. Statistically negative concentration
values are often observed. These measurements must be
included in the estimation of true mean values. For
this and other reasons, the frequency distributions of
measured values appear to be log-normal. Difficulties
in stating the true means and standard deviations are
discussed for these situations.
The Fort St. Vrain High Temperature Gas-cooled (HTGR) power

reactor, operated by Public Service Company of Colorado i s located
approximately MO miles north of Denver at the confluence of the
South Platte r i v e r and St. Vrain creek. I t i s the only gas-cooled
power reactor i n the country and while i t has had operating d i f f i -
c u l t i e s , the nuclear aspects of the design have great promise f o r
the following reasons:
1. Net e l e c t r i c a l e f f i c i e n c y i s 39.2$
2. Extremely low in-house worker radiation dose rates
3. Extremely low r a d i o a c t i v i t y release to the environment
The l a s t two c h a r a c t e r i s t i c s are primarily due to the unique HTGR

fuel element design and the rather innocuous environment of the
core. Table I i l l u s t r a t e s this point with a comparison of HTGR
effluents with those from B o i l i n g Water Reactors (BWR) and
Pressurized Water Reactors (PWR).
0097-6156/88/0361 -0266$06.00/0

14. JOHNSON & JOHNSON Radioactivity Analyses & Detection Limits 267
Table I: Radioactivity Released by Reactor Type
H-3 (Ci/MWe-year) Cs-137 (mCi/MWe-year)

Gaseous Liquid Liquid
BWR 0.05 0.1 25
PWR 0.2 1.2 1 .7
HTGR 0.1 6.5 2.8 E-4
Since the r a d i o a c t i v i t y release i s so low, the r a d i o a c t i v i t y con

centration i n a l l environmental sample types i s e s s e n t i a l l y a l l due
to natural radiation background sources with f a l l o u t and releases
from other industries p r i n c i p a l l y medical, superimposed upon i t .
The background source term due to primordial radionuclides i s
e s s e n t i a l l y constant and the frequency d i s t r i b u t i o n can be normally
or log-normally distributed. The f a l l o u t source term i s time de
pendent and highly variabl
Since the objective o
i s to document the presenc
to reactor effluents, i t i s imperative that background i n the
reactor environs be documented. For Fort St. Vrain the preoper
ational period occurred during periods of s i g n i f i c a n t Chinese
weapons test f a l l o u t and e s s e n t i a l l y a l l r a d i o a c t i v i t y data shows
lower values after reactor start-up. This makes comparison of
operational periods to preoperational periods of no value. Thus i t
i s necessary to compare s i t e data near the reactor or i n the
predominant wind directions to control data.
Figure 1 shows gross beta p a r t i c u l a t e concentrations i n a i r .
Gross beta measurements include both background (due to -0 and
the U-238 and Th-232 series) with f a l l o u t superimposed. Shown are
the half yearly arithmetic means f o r the four stations located i n
the predominant wind directions from the reactor and the means f o r
the reference (control) a i r sampling stations. There i s no
difference i n the means between the two station sets. The ex
tremely large temporal v a r i a t i o n i s due to f a l l o u t from the Chinese
atmospheric weapons tests during the period. (Chernobyl f a l l o u t
measured during the spring of 1986 reached weekly maxima of 560
3
fCi/m ).
Figure 2 shows that f a l l o u t Cs-137 concentrations i n milk are
log-normally distributed. (The geometric mean (median) multiplied
by and divided by the geometric standard deviation includes 68.1 %
of the area under the frequency d i s t r i b u t i o n . )
Figure 3 shows gross beta p a r t i c u l a t e concentrations i n pre
c i p i t a t i o n at the two f a c i l i t y s i t e c o l l e c t o r s . The t o t a l data
( s o l i d c i r c l e s ) , i f assumed to be log-normally distributed, would
be very misleading. There are c l e a r l y two separate frequency d i s
tributions with very d i f f e r e n t geometric means and geometric stand
ard deviations. The high a c t i v i t y concentration d i s t r i b u t i o n i s
assumed to be f a l l o u t and the lower one due to natural background.
Iodine-131 i s certainly the key radionuclide i n reactor
environmental monitoring. I t i s an indicator of release since the
radionuclide i s almost always gaseous and i t i s the predominant
f i s s i o n product contributor to radiation dose to the general public
due to i t s food chain mobility. Reactor license requirements
therefore put great emphasis on 1-131 measurements i n a i r and

Figure 1. Gross beta concentrations i n a i r for 1974 through

1985.

Figure 2. Cs-137 in milk, p C i / l i t e r , composites from adjacent

area, 197^ and 1975.

milk. The detection l i m i t (LLD) f o r 1-131 i n a i r required by the

3
USNRC i s 70 fCi/m . Using activated charcoal as the c o l l e c t i o n
medium and a fixed a i r sampler flow rate, i t i s possible to achieve
this value with Ge(Li) spectroscopy and counting times of
approximately 300 minutes. (The net count rate due to 1-131 i s
obtained by subtracting the sum of the count rates i n four channels
adjacent to the peak on each side from the sum of the count rates
i n the eight channels i n the peak.)
During a l l of 1985 the Fort St. Vrain reactor was shut down,
therefore no 1-131 could possibly have been released. 1-131 i s
released from l o c a l hospitals after patient diagnostic procedures
and treatment with 1-131 but this release i s confined to surface
waters. Figure H shows the measured 1-131 concentration at a l l
seven of the a i r sampling stations f o r a l l of 1985. I f , i n f a c t ,
there was no source term f o r 1-131 i n a i r , the expected frequency
d i s t r i b u t i o n of the measured value would b du onl t methodo
l o g i c a l uncertainty an
buted. The arithmetic
3
shows the observed mean was actually -1.5 fCi/m indicating a
negative systematic error, probably due to bias i n the subtraction
method used to obtain net count rate. The frequency d i s t r i b u t i o n
was indeed normal. This i l l u s t r a t e s the necessity of including a l l
negative values i n data analysis. I f only positive values were
averaged, the mean would obviously be biased toward a f a l s e higher
value. I t also i l l u s t r a t e s a method of determining systematic
error. Using the s i m p l i f i e d expression of currie 0_) f o r S ( S = Q Q
2.33 s^) the a - p r i o r i , S f o r our measurement parameter was 33

3
fCi/m . For the data shown i n Figure 4 (n =356), only 1.2$ of the
values exceeded the S value where 5% f a l s e positive values would
be expected. I f the negative bias i s taken into account and the
3
d i s t r i b u t i o n normalized to a mean of zero by adding 1.5 fCi/m ,
3
more of the values would be greater than 33 fCi/m bringing the
f a l s e positive percentage closer to the expected 5$.
The a - p r i o r i S determined from the sum of the count rates i n
c
channels adjacent to the peak can be compared to the S based on q
determined f o r the net peak count rates for the 356 t r i a l s :

7 2 3
= S..M. (/ * = (0.63) (356)^ = 12 fCi/m

3
S = 1.65 = (1 .65) (12) = 20 fCi/m
c
As stated above, detection of effluent releases depends upon
comparing mean a c t i v i t y concentrations over a period of time with
those i n a reference or control area, as close i n c h a r a c t e r i s t i c s
as possible to the reactor area. To properly compare means the
appropriate variances must be used. When reporting standard devi
ations therefore, i t i s important to remove the f r a c t i o n of the
t o t a l uncertainty which i s due to the method. Since the precision
(random uncertainty) to be attached to the method i s commonly
determined with spiked samples, only the c o e f f i c i e n t of v a r i a t i o n
may be used. Figures 5, 6 and 7 i l l u s t r a t e an approach to sub
t r a c t i n g the c o e f f i c i e n t of v a r i a t i o n of the method from Sr-90
measurements i n s o i l . Figure 5 shows that the observed t o t a l

CUMULATIVE %
Figure 3. Total gross beta i n p r e c i p i t a t i o n , F-1 and F-i
combined July 1976 through June 1977 (F-1 and F-4 were the
sample locations f o r two large precipitation collector
funnels). _
3
X = -1.5fCi/m
S.E.M. - 0.63
normal dist., = 0.01
= 356
3
Se = 33 fCi/m
40 25 10 0 5 20 35
3
fCi/m
Figure 4. 1-131 concentrations i n a i r , 1985.

2 2 2
= +
Total Env. Method
If Independent,
2 2 + 2 + 2 + . 2
^Method ^Sampling ^Sample Chemical Counting
Method Prep. Sep.
Since q 2
often determined with spiked sample,
method
Use:
x v x
V X /Total * ' Env. Method
Figure 5. Determination of uncertainty due to method and

environmental v a r i a t i o n .
From a single homogenized sample
= 10
(-2) = 0.22 (i) = 0.05

V X
'method method
This includes counting uncertainty
Method: 89.90 S r b y SrC0 i,SrN0 i 3 3
9 0 9 0
Sr by Y(OH) i 3
Sr chemical yield by AAS of carrier.
Figure 6. Precision (random uncertainty) of Sr-90 method,

1979.

14. JOHNSON & JOHNSON Radioactivity Analyses Detection Limits 273
= 38
X - 220 pCi/k
- 140 pCi/kg
Figure 7. Sr-90 measurements i n surface s o i l , 1980, F t . St.

Vrain environs.

274 D E T E C T I O N IN A N A L Y T I C A L C H E M I S T R Y
c o e f f i c i e n t of v a r i a t i o n term squared i s e q u a l t o the c o e f f i c i e n t

of variation of the method squared and the square of the
c o e f f i c i e n t of v a r i a t i o n due o n l y t o e n v i r o n m e n t a l f a c t o r s . Since
i t i s o n l y t h i s l a s t term t h a t s h o u l d be used t o d e s c r i b e the
r e s u l t s f o r a g i v e n environment, the u n c e r t a i n t y due t o the method
must be s u b t r a c t e d . I n t h i s case 10 r e p l i c a t e subsamples from a
homogenized large sample were analyzed identically and the
c o e f f i c i e n t of v a r i a t i o n determined t o be 0.22 ( F i g u r e 6 ) . T h i s i s
not l a r g e c o n s i d e r i n g the number of c h e m i c a l and c o u n t i n g steps
involved i n the method. F i g u r e 7 shows t h a t i f the method
u n c e r t a i n t y i s s u b t r a c t e d (as the c o e f f i c i e n t of v a r i a t i o n squared)
from the t o t a l c o e f f i c i e n t of v a r i a t i o n squared of a s e t of 38 s o i l
samples a n a l y z e d f o r S r - 9 0 , the r e s u l t i n g c o e f f i c i e n t of v a r i a t i o n
was 61$. A l t h o u g h i t i s t r u e t h a t o f t e n the m e t h o d o l o g i c a l
uncertainty i s small compared t o b i o l o g i c a l or environmental
v a r i a t i o n , the approach must s t i l l be used When the u n c e r t a i n t y
frequency d i s t r i b u t i o n
d i s t r i b u t i o n i s log-normal
I n c o n c l u s i o n , i t i s o b v i o u s t h a t e n v i r o n m e n t a l measurements
a t r a d i o a c t i v i t y l e v e l s n e a r , or i n the case of F o r t S t . V r a i n ,
below preoperational backgrounds present a complex problem.
A t t e n t i o n must be g i v e n t o d e t e r m i n a t i o n of the p r o p e r f r e q u e n c y
d i s t r i b u t i o n of the d a t a and t o the p r o p e r s t a t i s t i c a l t e s t t o
compare f a c i l i t y a r e a d a t a t o t h a t from background locations.
Statistical methodology cannot always be r i g o r o u s l y applied and
must be combined w i t h a common sense approach t o the available
data. N e g a t i v e v a l u e s must always be i n c l u d e d i n the c a l c u l a t i o n
of means and s t a n d a r d d e v i a t i o n s i n o r d e r t o a v o i d b i a s i n g r e s u l t s
towards h i g h v a l u e s .
L i t e r a t u r e Cited
1. C u r r i e , L. A. Anal. Chem. 1968, 40, 586-693.


Chapter 15
Detection Limits for Amino Acids

in Environmental Samples
1 2
P. E. Hare and P. A. St. John
1
Geophysical Laboratory, 2801 Upton Street, N.W., Washington, DC 20008
2
St. John Associates, Inc., 4805 Prince Georges Avenue,
Beltsville, MD 20705
Liquid chromatograph
derivatives and laser-induced fluorescence can separate
and detect sub-femtomole levels of amino acids. The
widespread distribution of amino acids and proteins
from living organisms produces nanomolar and higher
levels of amino acid material in most environments of
the earth's surface. Contamination usually occurs to
some extent during sample collecting and processing and
must be recognized and addressed before meaningful
amino acid concentrations and distributions can be
obtained from environmental samples.
Amino acids are distributed ubiquitously throughout much of the

earth's crust, including the atmosphere Their occurrence and
important r o l e i n l i v i n g organisms are well known, but amino acids
have also been found i n f o s s i l s and rocks hundreds of m i l l i o n s of
years and even b i l l i o n s of years old (2). Amino acids have been
reported at p a r t s - p e r - b i l l i o n l e v e l s i n e x t r a - t e r r e s t r i a l samples
such as the Apollo moon rocks as well as i n several meteorites (3,4).
Even d i s t i l l e d water, reagent-grade HC1 and other chemicals f r e -
quently contain trace amounts of amino acids. Figure 1 summarizes
the l e v e l s of amino acids found i n several samples of environmental
interest.
Current a n a l y t i c a l techniques for the analysis of the common
amino acids are able to detect less than femtomole (10~15) amounts
using l i q u i d chromatographic methods with fluorescent derivatives.
However, t h i s l e v e l of s e n s i t i v i t y may be extremely d i f f i c u l t to
u t i l i z e because of the widespread presence of amino acids from l i v i n g
organisms. For example, the r e l a t i v e l y high l e v e l s of amino acids i n
human body tissues and f l u i d s make human f i n g e r p r i n t s or even a
person's breath a potential serious contaminant i n detecting amino
acids i n s p e c i f i c environmental samples (5). Sample c o l l e c t i n g and
preparation are major concerns to amino acid trace analysis. Even
with c a r e f u l handling during sample c o l l e c t i o n and preparation i t i s
always possible that the sample may have been contaminated i n s i t u
before i t s c o l l e c t i o n .
0097-6156/88/0361-0275$06.00/0

HORN AMINO ACID ABUNDANCES

1 g A . A . / g S.
HAIR
BONE
10.-1k FINGERPRINT
-J
o.
< 10 r-2
MODERN SHELL (Mytilus)
cc
UJ
MODER
10r-3k
mg/g|
LU
FOSSIL SHELL (Mercenaria) MIOCENE
10" SOIL
METEORITE (Murchison)
>

10"
UJ
CC
CO
SHALLOW MARINE WATER
10- kPPM
Mg/g
GUNFLINT CHERT (Precambrian)

S DEEP MARINE WATER
<
u. 10-7

I- METEORITE (Allende)
X
UJ
10" APOLLO SAMPLES
u.

10 -9 IPPB
ng/g
10 - 1 0 I
F i g u r e 1. D i s t r i b u t i o n o f amino a c i d s i n s e l e c t e d components o f
t h e E a r t h ' s c r u s t . R a t i o o f w e i g h t o f r e c o v e r e d amino a c i d s t o
w e i g h t o f s a m p l e . P u r e p r o t e i n s p l o t a s 1 gram p e r gram.

15. H A R E & ST. J O H N Detection Limits for Amino Acids 277
Because of the widespread d i s t r i b u t i o n of l i v i n g organisms i t i s

often d i f f i c u l t to interpret the significance of finding amino acids
at trace l e v e l s (<10""6 g/g) i n geological samples. The presence or
absence of unstable amino acids such as serine and threonine as well
as D-amino acids from racemization can be helpful i n distinguishing
older i n s i t u f o s s i l amino acid assemblages from modern amino acid
contaminants. Stable isotope r a t i o s of nitrogen and carbon can also
be helpful i n interpreting sources and pathways for amino acids found
i n environmental samples. Micromole l e v e l s of the amino acids are
needed for isotope studies, and t h i s l e v e l may not be available for
some samples.
In spite of the d i f f i c u l t i e s encountered with contaminants from
l i v i n g organisms, several studies have shown the usefulness of deter-
mining the amino acid compositions and concentrations i n geological
samples ranging from nearly modern samples to samples b i l l i o n s of
years old (I). As technique
analysis are further developed
acids i n various environment increasingly helpfu g
pathways of amino acids and proteins from l i v i n g organisms into the
environment. Food chains as well as airborne and waterborne diseases
are research areas that should benefit from data on the pathways of
d i s t r i b u t i o n of amino acids and proteins throughout the crust of the
earth.
Approaches to Amino Acid Analysis by Liquid Chromatography
The analysis of amino acids by l i q u i d chromatography follows one of

three main approaches depending on how the amino acids or their
derivatives are chromatographed and detected. The f i r s t and most
straightforward method i s to separate and elute the free amino acids
from a column while monitoring the effluent at around 210 nm wave-
length.
The second approach involves post-column d e r i v a t i z a t i o n of the
amino acids, which improves the s e n s i t i v i t y for detection by creating
either a highly fluorescent derivative or one with a strong chromo-
phore. The t r a d i t i o n a l approach for amino acid analysis uses post-
column d e r i v a t i z a t i o n i n which the reagent ninhydrin mixes with the
column effluent and reacts with the amino acids to form a derivative
that can be detected by absorbance at 570 nm and 440 nm.
The t h i r d approach i s to make the amino acid derivatives before
i n j e c t i o n into the column. This pre-column d e r i v a t i z a t i o n method has
p o t e n t i a l l y the highest s e n s i t i v i t y and i s increasing i n popularity
i n a n a l y t i c a l laboratories. By choosing appropriate c h i r a l
( o p t i c a l l y active) reagents for d e r i v a t i z a t i o n and/or chromatography,
i t i s possible to make diasteroisomers that can be separated during
chromatography or diastereomeric interactions that allow determina-
tion of the o p t i c a l configuration of the amino acids i n the sample.
The d i s t r i b u t i o n of D and L amino acids among l i v i n g organises i s not
well known and a systematic study would be helpful i n interpreting
the geological and environmental occurrences* of D amino a c i d s .
UV Detection of Underivatized Amino Acids
A l l of the common amino acids, including hydroxyproline and p r o l i n e ,

show some absorption of UV l i g h t at around 200-210 nm (6). This non-

destructive detection method i s p a r t i c u l a r l y useful for a preparative

system for amino acids using ion-exchange resins and HC1 e l u t i o n .
Adequate separation i s achieved s t a r t i n g with an i s o c r a t i c (constant
composition) e l u t i o n of 0.6M HC1 followed by a l i n e a r gradient to
2.5M HC1 to elute phenylalanine and the basic amino acids: h i s t i d i n e ,
l y s i n e and arginine. Work on nitrogen and carbon stable isotopes i n
i n d i v i d u a l amino acids has used t h i s method of separation success-
f u l l y . Evaporation of the v o l a t i l e HC1 eluent f a c i l i t a t e s sample
preparation for isotope a n a l y s i s . Considerable i s o t o p i c f r a c t i o n a -
tion of the order of 1% occurs during chromatography, making i t
essential to achieve complete resolution and recovery of the entire
chromatographic peak. This method could be developed into a suitable
a n a l y t i c a l technique for amino acid analysis, with s e n s i t i v i t y i n the
low nanomole or even picomole l e v e l s . I t detects the secondary amino
acids proline and hydroxyproline and needs no pre-column or post-
column d e r i v a t i z a t i o n , whic
Post-Column Derivatizatio
Modern amino acid methods of analysis started with the paper by

Spackman, Stein and Moore i n 1958 (7). Chromatographic columns were
packed with ion-exchange resin p a r t i c l e s of uniform s i z e . Sodium
c i t r a t e buffers were pumped through the columns eluting the amino
acids. A 150 cm by 0.9 cm column separated the a c i d i c and neutral
amino acids, while a 15 cm by 0.9 cm column separated the basic amino
acids from a second sample a l i q u o t . Ninhydrin reagent was pumped
continuously into a tee and mixed with the column effluent containing
the separated amino acids. The mixture was pumped through a length
of t e f l o n tubing heated i n a b o i l i n g water bath for 15 minutes to
e f f e c t the reaction of the amino acids with the ninhydrin. The
mixture was monitored i n a flow-through dual-channel photometer at
570 and 440 nanometers. A multipoint recorder printed out a continu-
ous record of the photometer output. The l i m i t of detection (taken
as signal to noise r a t i o of 2) was around 50 nanomoles (10~9) for
most of the amino acids. The t o t a l analysis time was 24 hours.
In 1987, using a pre-column d e r i v a t i z a t i o n method, the common
amino acids can be analyzed i n less than 20 minutes with the l i m i t of
detection less than 0.5 femtomole (10~15) (S), This represents an
increase i n s e n s i t i v i t y of at least eight orders of magnitude and a
decrease i n analysis time by around two orders of magnitude! A b r i e f
description of how these changes have come about follows.
The f i r s t s i g n i f i c a n t improvement came as a result of using
smaller p a r t i c l e - s i z e resins to reduce band-spreading. This permits
the use of smaller volume columns, which require less volume of
mobile phase to elute each amino acid (9). This i n turn increases
the concentration of amino acids i n the column eluent. With better
resins and smaller volume columns alone, i t became possible to
improve the detection l i m i t s by two orders of magnitude. At the same
time i t was possible to shorten analysis times to around a few hours.
Another improvement was a f l o w c e l l design that increased s e n s i t i v i t y
by another order of magnitude by increasing the pathlength while
decreasing the volume of the f l o w c e l l . Further improvement i n s e n s i -
t i v i t y has resulted from the development of s t i l l smaller column
volumes with yet f i n e r size p a r t i c l e packings (5-micron resin beads),
and i n p a r t i c u l a r the development of fluorescent derivatives of amino
acids (10,11).

In post-column d e r i v a t i z a t i o n systems, the d e r i v a t i z a t i o n

reagent i s continuously mixed with the column e f f l u e n t ; consequently,
the presence of trace amounts of amino acids or i n t e r f e r i n g compo
nents such as ammonia i n the column eluent l i m i t s the ultimate sensi
t i v i t y attainable. Figure 2 shows the e f f e c t s of buffer contaminants
on the baseline of a t y p i c a l high s e n s i t i v i t y analysis (12). The
basic contaminants are held up on the column u n t i l e l u t i o n by the
f i n a l buffer. Detection of the basic amino acids below 10-picomole
l e v e l s i s d i f f i c u l t using t h i s approach. The use of a shorter second
column with i s o c r a t i c e l u t i o n for the analysis of j u s t the basic
amino acids i s a good solution to the problem but does require the
use of a second aliquot of sample.
Chromatography of Amino Acid Derivatives
Pre-column d e r i v a t i z a t i o
acid derivatives p o t e n t i a l l
able for amino acid analysis (<10~15 molar). The important advan
tage i n t h i s approach i s that trace amino acid contaminants i n the
column eluents do not i n t e r f e r e as i n the post-column systems. Only
amino acids i n the sample that have been derivatized are detected,
and consequently the baseline i s not influenced by possible contami
nants contained i n the eluents.
There i s a wide choice of amino acid derivatives that can be
used. Phenylisothiocyanate (PITC) reacts with both primary and
secondary amino groups to form moderately stable phenylthiocarbamyl
(PTC) derivatives that are separated on a reversed-phase column and
can be detected i n a UV detector at 254 nm wavelength (13). Detec
t i o n l i m i t s are at about the picomole l e v e l . Sample d e r i v a t i z a t i o n
with PITC takes around 20 minutes and requires close attention to
d e t a i l s i f consistent r e s u l t s are to be obtained. The presence of
s a l t s such as NaCl i n the sample i n t e r f e r e s with the d e r i v a t i z a t i o n
of several of the amino acids, and care must be taken i n processing
samples containing s a l t s .
The development of fluorescent derivatives of amino acids and
t h e i r chromatography on reversed-phase columns y i e l d a s i g n i f i c a n t
gain i n s e n s i t i v i t y . Many fluorescent derivatives of amino acids are
a v a i l a b l e that greatly enhance the s e n s i t i v i t y of detection. 0-
phthaldialdehyde/mercaptoethanol () reagent reacts with most of
the common amino acids (but not proline) to form fluorescent deriva
tives (14). Because derivatives are not very stable, i t i s
e s s e n t i a l to chromatograph the derivatives within a few minutes.
In order to achieve consistent a n a l y t i c a l r e s u l t s , i t i s necessary to
automate or to time accurately the d e r i v a t i z a t i o n step. Amino acid
analysis with pre-column takes l e s s than 15 minutes including the
d e r i v a t i z a t i o n step. I t has become a popular technique wherever
proline values are not necessary.
The search for improved pre-column derivatives for amino acid
analysis i s a continuing process. Dansyl derivatives are stable and
have been used for amino acid analysis, but hydrolysis products
(dansyl OH) of the dansyl reagent are d i f f i c u l t to eliminate and can
i n t e r f e r e with the dansyl amino acid peaks (15). FM0C-C1 ( 9 - f l u o r -
enylraethyl chloroformate) i s another derivative that has been used
and that shows promise for the analysis of the secondary amino acids
proline and hydroxyproline as well as the primary amino acids (16).

A. AMINO ACID S T A N D A R D
B. H U M A N FINGERPRINT (WET)
I 1 1 I I 1
10 15 20 25
MINUTES
F i g u r e 2. Post-column amino a c i d chromatograms showing l i m i t a t i o n s

to d e t e c t i o n l i m i t s due t o contaminants i n the m o b i l e phase o f the
l i q u i d chromatography system. Ion-exchange column w i t h sodium
c i t r a t e b u f f e r s and r e a g e n t f o r post-column r e a c t i o n . V e r t i c a l
a x i s i s f l u o r e s c e n c e w i t h the same s e n s i t i v i t y f o r each r u n . A.
Amino a c i d s t a n d a r d w i t h 100 p i c o m o l e s of each amino a c i d i n j e c t e d .
B. Human f i n g e r p r i n t showing h i g h l e v e l s of s e r i n e , g l y c i n e , and
ornithine. C o n t i n u e d on next page.

C. B A S E L I N E B L A N K
10 15

20 25
MINUTES
Figure 2 . C o n t i n u e d . C. No sample i n j e c t e d . M o b i l e phase con

t a m i n a n t s accumulate on the column and e l u t e w i t h change i n m o b i l e
phase.

The i d e a l pre-column derivative for amino acids, when i t i s found,

should react with both secondary and primary amino acids and also
make i t possible to separate enantiomers during chromatography.
Detection Using Laser-Induced Fluorescence (LIF)
Fluorescence intensity i s d i r e c t l y proportional to e x c i t a t i o n l i g h t

i n t e n s i t y . The use of laser e x c i t a t i o n would seem to be a l o g i c a l
way to improve detection l i m i t s . The p o t e n t i a l i s great: monochro
matic r a d i a t i o n , sharply collimated beam geometry, and high intensity
at the laser wavelengths. These features match well with the
requirements for fluorometric l i q u i d chromatography detectors, i n
p a r t i c u l a r the c o n f l i c t i n g requirements for s m a l l - c e l l volumes, long
o p t i c a l path lengths, and e f f i c i e n t minimization of stray l i g h t .
Unfortunately, several aspects of laser l i g h t sources l i m i t t h e i r
u t i l i t y as fluorescence e x c i t a t i o Stabilit f output
intensity i s poor, and nois
continuous wave and pulse
sources. Wavelength a v a i l a b i l i t y i s limited to a few choices and i s
very limited i n the u l t r a v i o l e t region of the spectrum. Wavelength-
tunable lasers are a v a i l a b l e , but the cost and complexity of these
systems poses a s i g n i f i c a n t barrier to t h e i r routine use i n
fluorescence detectors.
Complexity and expense notwithstanding, laser e x c i t a t i o n has
been shown to produce good detection l i m i t s for amino acid analysis.
Roach and Harmony (8) have demonstrated detection l i m i t s of 5 to 15
femtomoles of hydrolysate amino acids i n standard solutions using 0-
phthaldialdehyde pre-column d e r i v a t i z a t i o n and 334-363 nm argon-ion
laser e x c i t a t i o n . These l i m i t s are on the order of tenfold better
than most commercial fluorescence detectors under s i m i l a r conditions.
These l i m i t s are lowered tenfold to 0.2 to 0.5 femtomoles using 2,3
naphthalenedialdehyde as the pre-column d e r i v a t i z i n g agent when a
457.9 nm argon laser l i n e i s used as the e x c i t a t i o n source. The
better detection l i m i t s using v i s i b l e - l i g h t e x c i t a t i o n are attributed
to the lower inherent noise l e v e l of the v i s i b l e - l a s e r l i n e and the
avoidance of high background fluorescence induced by u l t r a v i o l e t
e x c i t a t i o n . An equally important contribution to the achievement of
a low dectection l i m i t appears to be due to the use of the flow c e l l
described by Sepaniak and Yeung (7). I t u t i l i z e s a quartz f i b e r
l i g h t pipe as an e f f i c i e n t o p t i c a l couple to a small-volume c e l l .
Novotny (18) has pointed out an alternate trend i n the e f f o r t to
u t i l i z e the potential power of the laser for HPLC detection, namely
the "tuning" of the chemistry of the system to match the a v a i l a b l e
laser wavelengths (19). The use of naphthalenedialdehyde as
discussed above i s a good example. An even more s t r i k i n g example i s
the work of Sauda, Imasaka and Ishibashi (20), who used indocyanine
green as a fluorescent tag for protein l a b e l i n g . The dye/protein
complex absorbs at about 765 nm and emits at 820 nm. This couples
well with the output of a pulsed semiconductor laser diode (780 nm,
15 mW). The r e s u l t i n g equipment i s capable of detecting 90 nanograms
of bovine serum albumin. The authors stress that blank fluorescence
i s "completely n e g l i g i b l e " i n near-infrared fluorometry; conse
quently, the low blank i s the dominant factor i n such low l i m i t s of
detection. These authors quite candidly point out that such
detection l i m i t s are of l i t t l e p r a c t i c a l use i n the absence of
reagent s p e c i f i c i t y .

Chemiluminescence Detection
Cherailuminescence detection of amino acids i s another promising

approach that has the potential for very high s e n s i t i v i t y . Replacing
electromagnetic excitation of analyte molecules with chemical
excitation eliminates the scattered l i g h t component of the usual
blank signal associated with a l i g h t source. Furthermore, the
s p e c i f i c i t y requirements of the chemi- or bioluminescence detection
reactions should reduce the o v e r a l l l e v e l of the blank resulting from
chemical contaminants.
Chemiluminescence r e s u l t s from energy transfer from peroxyoxa-
l a t e reaction systems to fluorescent analytes. This approach to high
s e n s i t i v i t y amino acid analysis was taken by Kobayashi and Imai (21)
and Miyaguchi et a l (22). Dansylated amino acids were detected i n
t h i s manner using bis (2,4,6-trichlorophenyl) oxalate and H2O2 as the
detection reagent. The
reviewed by Imai and Weinberge
well-known luminol-peroxid syste pre
sented by MacDonald and Nieman (24). They chose a quenching mode of
operation whereby the Co(II)-enhanced luminol/peroxide luminescent
solution was continuously pumped and mixed with the effluent of the
HPLC column. Complexation of the Co(II) by analyte amino acids
e f f e c t i v e l y removes the Co(II) from i n t e r a c t i o n i n the chemilumi-
nescence reaction sequence and r e s u l t s i n decreased l i g h t emission.
This type of procedure i s dependent on the formation constants of the
amino acid-Co(II) complexes. Limits of detection range from about 4
picomoles for good complex formers such as h i s t i d i n e and cystine to
as high as 20 nanomoles f o r poor complexers such as aspartic acid.
These r e s u l t s are not d i r e c t l y comparable with the detection l i m i t s
quoted f o r the peroxyoxalate procedure because of the r e l a t i v e l y
i n e f f i c i e n t bonded cation exchange column used to separate the amino
acids.
Discussion
I t i s possible to detect sub-femtomole l e v e l s of amino acids and

t h e i r enantioraers with state-of-the-art l i q u i d chromatography
systems. Under most circumstances, these l e v e l s of detection are
unattainable due to the presence of amino acids and proteins i n
v i r t u a l l y a l l environments. In order to r e a l i z e the potential l i m i t s
of detection, i t i s necessary to u t i l i z e new approaches to sample and
reagent preparation. To approach the l i m i t s of detection possible
with the technique, a properly used clean room f a c i l i t y i s a must. A
serious and seldom recognized source of contamination i n amino acid
chemistry i s the use of 6N HC1 for the hydrolysis of peptides and
proteins to free amino acids. Even reagent-grade HC1 frequently
contains amino acid contaminants, but even more serious i s the
storage and handling of the reagent. Whenever the HC1 reagent bottle
i s opened, the HC1 can become seriously contaminated by airborne
contaminants such as dust and pollen. Laminar flow hoods and special
handling reduce the problem but do not eliminate i t . Vapor phase
hydrolysis with 6M HC1 i s another promising approach to reducing
contaminant l e v e l s . Analysis even at picoraole and low nanoraole
l e v e l s needs careful controls during sample preparation. Controls
should include samples heated to remove possible amino acids as well

as frequent reagent blanks processed and analyzed to determine

background contaminant l e v e l s . It i s embarrassing to find out that
the sample contained the same l e v e l s of amino acids as the reagent
control blanks!
In geological samples the problem of i n s i t u contamination by
l i v i n g organisms and t h e i r metabolic products l i m i t s the usefulness
of amino acid geochemistry i n older geological samples. The o r i g i n a l
l e v e l s of proteins i n s h e l l s , bones and teeth decrease with time due
to chemical breakdown and leaching. Ground water and other sources
contain l e v e l s of amino acids that can sometimes exceed the l e v e l s of
amino acids l e f t from the o r i g i n a l organic matter i n the f o s s i l .
When t h i s happens, amino acids can diffuse into the f o s s i l and show
higher than expected l e v e l s .
This frequently encountered s i t u a t i o n i s usually recognized by
the presence of excessive l e v e l s of serine. Serine i s one of the
least stable amino acid
rapidly i n progressivel
in contaminants are generally L-configuration, agai pointing
to l i v i n g organisms as the source. The o r i g i n a l protein i n f o s s i l s
undergoes hydrolysis and racemization with increasing l e v e l s of D-
amino a c i d s . It may not always be possible to d i s t i n g u i s h
racemization-derived D-amino acids from c e l l - w a l l D-amino acids i n
l i v i n g bacteria.
In the e a r l i e r days of amino acid geochemistry i t was necessary
to extract the amino acids from r e l a t i v e l y large samples, frequently
entire s h e l l s or bones. As l i m i t s of detection have improved i t has
been possible to study the d i s t r i b u t i o n of amino acids within a
single s h e l l and even detect gradients from the surface to the
i n t e r i o r of a s h e l l or bone. Frequently the outside of a contami
nated f o s s i l w i l l show much higher l e v e l s than the i n t e r i o r as well
as a d i f f e r e n t composition. An uncontaminated f o s s i l shows less
concentration of amino acids on i t s surface and a greater concentra
t i o n i n i t s i n t e r i o r . Only with improved l i m i t s of detection have
studies on d i s t r i b u t i o n of amino acids i n single fossilsand even i n
microfossils such as diatoms and foramsbeen f e a s i b l e .
With attention to sample c o l l e c t i n g and handling i t w i l l be
possible to study e f f e c t i v e l y the occurrences and d i s t r i b u t i o n of
amino acids throughout the crust of the earth and even i n regions
(submarine vents, volcanic areas and e x t r a t e r r e s t r i a l regions, e t c . )
where a b i o t i c synthesis of amino acids has been suggested. Prospects
are e x c i t i n g !
Literature Cited
1. Hare, P. E. In Organic Geochemistry, Methods and Results;

Eglinton, G.; Murphy, M.T.J., Eds.; Springer-Verlag: New York,
1969; p 438.
2. Schopf, J . W.; Kvenvolden, . .; Berghoorn, E. S. Proc. Natl.
Acad. Sci. 1968, 59, 639-648.
3. Kvenvolden, . .; Lawless, J . G.; Ponnaperuma, C. Proc. of
Natl. Acad. Sci. 1971, 68, 486.
4. Harada, K.; Hare, P. E. In Biogeochemistry of Amino Acids;
Hare, P. E . ; Hoering, T. C.; King, K., Eds.; John Wiley & Sons,
Inc.: New York, 1980; p 169.

15. HARE & ST. JOHN Detection Limits for Amino Acids 285
5. Hamilton, P. B. Nature 1965, 205, 284.

6. Schuster, R. Anal. Chem. 1980, 53, 617.
7. Spackman, D.; Stein, W. H.; Moore, S. Anal. Chem. 1958, 30,
1190.
8. Roach, M. C.; Harmony, M. D. Anal. Chem. 1987, 59, 411.
9. Hamilton, P. B. Anal. Chem. 1963, 35, 2055.
10. Hare, P. E. Methods in Enzymol. 1977, 47, 3.
11. Benson, J . R.; Hare, P. E. Proc. Natl. Acad. Sci. 1975, 72,
619.
12. Hare, P. E. In Chemistry and Biochemistry of the Amino Acids;
Barrett, G. C., Ed.; Chapman and Hall: New York, 1985; p 415.
13. Bidlingmeyer, . .; Cohen, S. .; Tarvin, T. L. J . Chromatogr.
1984, 336, 93-104.
14. Hill, D. W.; Walters, F. H.; Wilson, T. D.; Stuart, J . D. Anal.
Chem. 1979, 51, 1338.
15. Perrett, D.; In Chemistry and Biochemistry of the Amino Acids;
Barrett, G. C., Ed.
16. Einarsson, S.; Josefsson
1983, 282, 609-618.
17. Sepaniak, M. J.; Yeung, E. S. J . Chromatogr. 1980, 190, 377-83.
18. Novotny, M. V. Chrom. Forum 1986, 1, 19-25.
19. Gluckman, J . C.; Shelly, D.; Novotny, M. J . Chromatogr. 1984,
317, 443.
20. Sauda, K.; Imasaka, T.; Ishibashi, N. Anal. Chem. 1986, 58,
2649-53.
21. Kobayashi, S.; Imai, K. Anal. Chem. 1980, 52, 424-27.
22. Miyaguchi, K.; Honda, K.; Imai, K. J . Chromatogr. 1984, 316,
501-05.
23. Imai, K.; Weinberger, R. Trends Anal. Chem. 1985, 4, 170-75.
24. MacDonald, .; Nieman, T. A. Anal. Chem. 1985, 57, 936-40.
RECEIVED March 2, 1987

Chapter 16
Real-World Limitations to Detection
1 2 3 4
David A. Kurtz , John K. Taylor , Larry Sturdivan , Warren B. Crummett ,
5 6 6
Charles R. Midkiff, Jr. , Robert L. Watters, Jr. , Laura J. Wood ,
7 8
W. William Hanneman , and William Horwitz
1
Department of Pathology, Pennsylvania State University,
University Park, PA 16802
2
3
CBM Branch, Chemical Research and Development
Chemical Research, Aberdeen Proving Ground, MD 21010
4
Analytical Laboratories, Dow Chemical Company, Midland, MI 48667
5
National Laboratory Center, Bureau of Alcohol, Tobacco, and Firearms,
Rockville, MD 20850
6
7
Center for Technology, Kaiser Aluminum and Chemical Corporation,
6177 Sunol Boulevard, Pleasanton, CA 94566
8
Food and Drug Administration, Washington, DC 20204
Data obtained at or near the limit of detection is

often improperly prepared by scientists and improper-
l y interpreted by the receiving public. This panel
offers advice to scientists preparing data to have
s t a t i s t i c a l support for the data produced. Four
examples from the real world illustrate interpreta-
tion problems: limit of detection reporting, the
r o l e of matrix effects and matrix v a r i a b i l i t y i n
forensic work, the effects of the blank in ICP work,
and federal regulatory problems.
Transferring c a r e f u l l y prepared data from the laboratory to the
general public can result i n problems and d i f f i c u l t i e s often not
considered beforehand. Data are created by analysts, usually at the
request of others. These data are of varying quality but are used i n
many areas of l i f e . Though the general purpose i s to solve problems,
problems are often created because of the varying q u a l i t i e s of the
data and the varying backgrounds of the users of the data. We have
convened a panel to discuss some of these problems suggesting i n some
cases solutions that could a l l e v i a t e them as problems.
There are four groups of people who use a n a l y t i c a l information
in daily l i f e . These are the non-experts (everyone except the
analysts), the governmental regulators and those regulated, the
rulers of the courts, and the analysts themselves. Each of these has
0097-6156/88/0361-0288$08.25/0

16. KURTZ ET AL. Real- World Limitations to Detection 289
d i f f e r e n t backgrounds, d i f f e r e n t areas of t h o u g h t , and d i f f e r e n t

purposes i n u s i n g a n a l y t i c a l i n f o r m a t i o n . The i n t e r a c t i o n between
t h e s e g r o u p s on s u c h i s s u e s c a n l e a d t o d i s c u s s i o n , f e a r , and
t h r e a t s . T h i s panel w i l l show how t h i s can come about.
The w r i t e r s each address s p e c i f i c areas where t h e r e have been
problems when d i v e r s e people a r e u s i n g a n a l y t i c a l i n f o r m a t i o n . Dr.
T a y l o r a d d r e s s e s the a n a l y s t s because t h i s i s of the f i r s t impor-
tance. Good i n f o r m a t i o n must come f i r s t . Dr. S t u r d i v a n p i c k s up
t h i s theme i n f u r t h e r d i s c u s s i n g how to o b t a i n good i n f o r m a t i o n i n
the narrower f i e l d of c a l i b r a t i n g i n s t r u m e n t a l response t o amounts or
c o n c e n t r a t i o n s i n samples. When a n a l y t i c a l r e p o r t s a r e made p u b l i c ,
m i s u n d e r s t a n d i n g s may o c c u r amongst the " n o n - e x p e r t s " as o u t l i n e d by
Dr. Crummett. He d i s c u s s e s problems i n such a r e a s as the presence of
d i o x i n i n human f a t and c o n t a m i n a t i o n of water s u p p l i e s w i t h t o x i c
solvents. Dr. Crummett d i s c u s s e s the r i s k s of f a l s e n e g a t i v e s and
f a l s e p o s i t i v e s . The i n t e r p r e t a t i o
d e t e c t i o n when t a k i n g th
available. Accordingly, , perhap
s i m p l i f i e d form, t o a c q u a i n t the reader w i t h the term, l i m i t of
detection. F u r t h e r r e a l - w o r l d problems i n the f o r e n s i c f i e l d are
d i s c u s s e d by Mr. M i d k i f f . He d e s c r i b e s some examples a t the l e g a l
i n t e r f a c e . In t h i s a r e a the a n a l y s e s a r e even more d i f f i c u l t because
o f the u n u s u a l m a t r i c e s found and the l a c k of sample t o work w i t h .
The governments, of c o u r s e , are asked by the people t o become a c t i v e
i n t h i s a r e a because of t h e i r concerns and l a c k of knowledge of b o t h
the a r e a s of a n a l y s i s and t o x i c o l o g y . Dr. Hanneman has shown t h a t
b r o a d l y w r i t t e n r e g u l a t i o n s can be d i f f i c u l t t o comply w i t h . Here i s
a l s o an example where r e s e a r c h e r s have not read the F e d e r a l R e g i s t e r ,
the c e n t r a l r e g u l a t i o n document, c a r e f u l l y enough t o get i t s t r u e
meaning. Hence, they can get the wrong i m p r e s s i o n and s u p p o r t an
e r r o n e o u s p l a n of a c t i o n . F i n a l l y , Dr. H o r w i t z takes the o t h e r s i d e
of the c o i n and defends the government i n p u t t i n g out good documenta-
t i o n about t h e i r w e l l planned r e g u l a t i o n s . He a l s o recommends some
b a s i c p o i n t s f o r a n a l y s t s t o remember as r u l e s o f t h e thumb t o
f o l l o w . . . a s l o n g as t h e y f i r s t know and f o l l o w good a n a l y t i c a l
practice.
The f o l l o w i n g t e x t d i s c u s s e s problem a r e a s . Here, a n a l y s t s are
w o r k i n g i n t r a c e amounts o r a t the l i m i t of d e t e c t i o n . Frequently,
problems o c c u r due t o the i n t e r a c t i o n between the a n a l y s t s , c o u r t s ,
and n o n - e x p e r t s . Key q u e s t i o n s t h a t might be answered a r e :
*** Can a q u a l i t y l a b e l be put on data?
*** How can we t e a c h the p u b l i c the concept o f
error?
*** Can we d i s c u s s d a t a , t a k i n g e r r o r i n t o a c c o u n t ,
w i t h o u t i n d u c i n g an a l a r m i s t r e a c t i o n ?
*** How do we u n d e r s t a n d the s p e c i a l d i f f i c u l t i e s
of o b t a i n i n g a n a l y s i s d a t a r e l a t i n g t o f o r e n s i c
problems
*** Can we i n t e r a c t w i t h f e d e r a l r e g u l a t o r s i n the
making of s u i t a b l e r e g u l a t i o n s ?
B a s i c Data Q u a l i t y ( D r . J . K. Taylor)
Have we always had i n the l a b o r a t o r y a s t a t e of s t a t i s t i c a l control

on a g i v e n problem? Dr. T a y l o r says i n t h i s r e g a r d , " S t a t i s t i c a l

c o n t r o l of the measurement p r o c e s s i s the f i r s t requirement f o r mean-
i n g f u l d e t e c t i o n , s i n c e w i t h o u t i t a l l d a t a are m e a n i n g l e s s . This
means t h a t a l i m i t i n g mean must be r e a l i z e d and a s t a b l e s t a n d a r d
d e v i a t i o n must be a t t a i n e d and m a i n t a i n e d . " By l i m i t i n g mean we have
i n mind a mean t h a t approaches the t r u e mean as the sample s i z e
becomes i n f i n i t e . A s t a b l e s t a n d a r d d e v i a t i o n means one t h a t does
not change i n s i z e w i t h the a d d i t i o n of a d d i t i o n a l samples. In o t h e r
words an a n a l y s t must be a b l e to reproduce h i s r e s u l t s a t a l l t i m e s .
O t h e r w i s e , any r e s u l t a t any time i s m e a n i n g l e s s .
" S t a t i s t i c a l c o n t r o l a p p l i e s to a l l p a r t s of the a n a l y t i c a l
s y s t e m - s a m p l i n g p r o c e s s , t h e c a l i b r a t i o n , t h e b l a n k , and the
measurement. S t a t i s t i c a l c o n t r o l i s a t t a i n e d by the q u a l i t y c o n t r o l
of the e n t i r e system and i n v o l v e s maintenance of r e a l i s t i c t o l e r a n c e s
for a l l c r i t i c a l operations. A system of c o n t r o l c h a r t s i s the b e s t
way to demonstrate a t t a i n m e n t f s t a t i s t i c a l c o n t r o l d t evaluat
the a p p r o p r i a t e standar
r e s u l t s of measurement
p e r i o d of t i m e , are p l o t t e d . S t a t i s t i c a l c o n t r o l i s demonstrated
when the v a l u e s are randomly d i s t r i b u t e d around t h e i r average v a l u e . "
C o n t r o l l i m i t s are o f t e n taken as + 2 or 3 s t a n d a r d d e v i a t i o n u n i t s
of these r e p l i c a t e s . Dr. T a y l o r a l s o adds, "Even the ranges of
d u p l i c a t e measurements of the a c t u a l samples t e s t e d can be p l o t t e d i n
a s i m i l a r manner to demonstrate a s t a b l e s t a n d a r d d e v i a t i o n . In
e i t h e r c a s e , t h e s t a t i s t i c s o f the c o n t r o l c h a r t s a r e t h e b e s t
d e s c r i p t o r s of the v a r i a b i l i t y of the measurement p r o c e s s . "
D r . T a y l o r c o n t i n u e s h i s g e n e r a l a d v i c e : " L a t e r speakers w i l l
d i s c u s s how measurement v a r i a b i l i t y q u a n t i t a t i v e l y d e f i n e s the l i m i t s
of d e t e c t i o n and q u a n t i f i c a t i o n . Due to the n a t u r e of measurement,
each l a b o r a t o r y ( a n a l y s t ) w i l l have somewhat d i f f e r e n t measurement
u n c e r t a i n t i e s , and hence d i f f e r e n t l i m i t s of d e t e c t i o n . Published
v a l u e s o f LOD's ( L i m i t s Of D e t e c t i o n ) , MDL's ( M e t h o d D e t e c t i o n
L i m i t s ) , or what-have-you f o r methodology are t y p i c a l , a t b e s t ; hence
they have no p r e d i c t a b l e q u a n t i t a t i v e r e l a t i o n to those o b t a i n e d by
any l a b o r a t o r y or a n a l y s t . Each must e v a l u a t e them f o r i t s e l f and
w i l l make somewhat d i f f e r e n t d e c i s i o n s c o n c e r n i n g p r e c i s i o n and
d e t e c t i o n when a n a l y z i n g the same samples. Each has the p r o f e s s i o n a l
o b l i g a t i o n to o b t a i n a l l i n f o r m a t i o n n e c e s s a r y to support the q u a l i t y
of i t s d a t a , which must be t e c h n i c a l l y sound and d e f e n s i b l e .
"Measurement u n c e r t a i n t y becomes c r i t i c a l l y i m p o r t a n t as i t
i n f l u e n c e s the d e c i s i o n p r o c e s s . In h i g h - a c c u r a c y a n a l y s i s , t h i s
i n f l u e n c e s the l a s t s i g n i f i c a n t f i g u r e . In t r a c e a n a l y s i s , i t
i n f l u e n c e s the f i r s t ( w h i c h i s a l s o the l a s t ) s i g n i f i c a n t f i g u r e .
No f i g u r e can be s i g n i f i c a n t w i t h o u t s t a t i s t i c a l s u p p o r t . Luckily,
many measurements are made w i t h more s i g n i f i c a n t f i g u r e s than are
needed f o r a d e c i s i o n p r o c e s s . A n a l y s t s must remember t h a t they
cannot use s t a t i s t i c s o b t a i n e d by any one e l s e t o support t h e i r own
d a t a and hence i n making d e c i s i o n s u s i n g i t .
"A f a c t t h a t i s n o t a l w a y s a p p r e c i a t e d i s t h a t t h e use o f
a p p r o p r i a t e methodology i s n e c e s s a r y but not s u f f i c i e n t f o r r e l i a b l e
measurement d a t a and/or f o r a t t a i n i n g s p e c i f i e d l i m i t s of d e t e c t i o n .
In the EPA Love Canal s t u d i e s , f o r example, the c o n t r a c t l a b o r a t o r i e s
v a r i e d by more than a f a c t o r of ten i n t h e i r d e t e c t i o n l i m i t s w h i l e
u t i l i z i n g the same methodology. A c c o r d i n g l y , i t i s f a l l a c i o u s to

a s s i g n q u a l i t y l a b e l s to d a t a based on t y p i c a l or even c o l l a b o r a t i v e
test r e s u l t s . Data from v a r i o u s l a b o r a t o r i e s may be of b e t t e r or
p o o r e r q u a l i t y than such i n d i c a t o r s . Only r e l i a b l e e s t i m a t e s based
on the performance of the producer should be u s e d .
"The wide d i f f e r e n c e s t h a t can o c c u r i n the q u a l i t y of d a t a have
serious i m p l i c a t i o n s for data compilations. U n l e s s t h e r e i s some way
to code and/or to a s s o c i a t e d a t a w i t h i t s u n c e r t a i n t y , poor d a t a can
be unduly i n f l u e n t i a l i n subsequent d a t a a n a l y s i s . In d e t e c t i o n
s i t u a t i o n s i m p r e c i s e d a t a l e a d s to l a r g e d e t e c t i o n l i m i t s ; hence
non-detection can have a d i f f e r e n t meaning, depending on how the
measurement was made.
1
" I n t h i s s p e a k e r s o p i n i o n the p r e c i s i o n a t t a i n e d i n measuring
r e a l samples [ed.: as opposed to measurement of s t a n d a r d s o n l y or
s t a n d a r d s d i s s o l v e d i n s u b s t i t u t e background m a t r i c e s ] i s the o n l y
r e l i a b l e b a s i s f o r d e c i s i o n s on d e t e c t i o n . I n l a r g e measurement
programs, the use of d u p l i c a t e - s a m p l control chart i th t
f e a s i b l e way to e s t a b l i s
defend l i m i t s of d e t e c t i o n
r e p l i c a t e measurements must be made on the samples t e s t e d f o r t h i s
purpose. Without documented d e m o n s t r a t i o n of p r e c i s i o n , the d a t a are
meaningless."
C a l i b r a t i o n E r r o r s (Dr. L. Sturdivan)
Another a s p e c t of d a t a q u a l i t y t h a t needs to be addressed i s the

e r r o r t h a t i s produced d u r i n g the c a l i b r a t i o n s t e p i n a n a l y s i s .
W h i l e , s t r i c t l y s p e a k i n g , c a l i b r a t i o n e r r o r s are s l i g h t l y o u t s i d e the
r e a l m of d e t e c t i o n l i m i t e r r o r s , i t i s a l s o t r u e t h a t much a n a l y t i c a l
work i s done a t the d e t e c t i o n l i m i t and the c a l i b r a t i o n e r r o r s t h a t
are produced can be p r o p o r t i o n a t e l y much l a r g e r a t the d e t e c t i o n
l i m i t than elsewhere on the g r a p h . Hence, i t i s most a p p r o p r i a t e to
d i s c u s s the m a t t e r at t h i s t i m e .
Most q u a n t i t a t i v e a n a l y t i c a l d a t a r e s u l t s from a c a l i b r a t i o n
step. W h e t h e r t h e d a t a comes f r o m a gas c h r o m a t o g r a p h , l i q u i d
chromatograph or from a s p e c t r o m e t e r , c a l i b r a t i o n i s r e q u i r e d .
C a l i b r a t i o n r e l a t e s i n s t r u m e n t a l response w i t h a s p e c i f i e d amount of
substance to be measured. C a l i b r a t i o n r e q u i r e s s t a n d a r d s to cover
and enclose the entire concentration range of interest. In
p r e p a r a t i o n of these s t a n d a r d s , however, Dr. S t u r d i v a n comments:
"Often the s t a n d a r d samples are made by s e r i a l d i l u t i o n . They're
t r e a t e d as though they are independent samples," w h i c h they a r e n ' t i n
t h a t c a s e . "Now t h e r e i s n o t h i n g wrong w i t h making s e r i a l d i l u t i o n s ,
but i f t h a t i s done, i t would be d e s i r a b l e to make two s e r i e s of
s e r i a l d i l u t i o n s , s i d e by s i d e . " What he i s r e f e r r i n g to comes from
the f o l l o w i n g d i s c u s s i o n : The b e s t method f o r the p r e p a r a t i o n of
s t a n d a r d s i s to prepare each c o n c e n t r a t i o n d i r e c t l y ( o r w i t h s e r i a l
d i l u t i o n s ) from neat (100% pure) compounds ( w i t h the same number of
d i l u t i o n s t e p s ) . This can be c o s t l y i n b o t h time and m a t e r i a l s . The
next b e s t method, and one t h a t can be shown to be s t a t i s t i c a l l y
n e a r l y c o r r e c t i s to s t a r t from two d i f f e r e n t neat s t a n d a r d s and
d i l u t e them i n p a r a l l e l t r e e - l i k e s t e p s . Tree s t r u c t u r e d i l u t i o n s
would be to p r e p a r e AA and AB from A; then p r e p a r e AAA, AAB, and AAC
from AA and ABA, ABB, and ABC from AB, e t c .

A n o t h e r approach i n sample p r e p a r a t i o n , and one h a v i n g some

b i a s , i s t o s t a r t from one neat w e i g h i n g measurement a s s u r i n g t h a t
each s t a n d a r d has the same number o f d i l u t i o n s t e p s i n i t s p r e p a r a -
tion. The d i l u t i o n s t e p s would be o f a t r e e s t r u c t u r e . Thus each
d i l u t i o n s t a n d a r d w i l l have t h e same p r e p a r a t i o n v a r i a n c e . The most
common s t a n d a r d s p r e p a r a t i o n method i s one where each s t a n d a r d i s
s e r i a l l y produced from t h e p r o c e e d i n g one. I n t h i s case a l a r g e b i a s
i s developed as d i l u t i o n s p r o c e e d . The e r r o r v a r i a n c e of d i l u t i o n
g e t s l a r g e r and l a r g e r w i t h each d i l u t i o n s t e p .
D r . S t u r d i v a n c o n t i n u e s , "There may be d i l u t i o n e r r o r s . These
tend t o o c c u r more f r e q u e n t l y a t the lower end, down t h e r e where t h e
d e t e c t i o n l i m i t i s . I f t h e r e i s a d s o r p t i o n t o the w a l l s o f the
c o n t a i n e r , the l o w e s t s t a n d a r d s resemble b l a n k s . "
Dr. S t u r d i v a n emphasizes "Good c o n f i d e n c e l i m i t s around t h e
c a l i b r a t i o n graph must be done i n e q u a l v a r i a n c e space" What he i s
s a y i n g i s the f o l l o w i n g : C a l i b r a t i o n s t a n d a r d s a r e prepared a t
various concentration level
The square o f the s t a n d a r
must be a p p r o x i m a t e l y e q u a l . A q u i c k l o o k a t any chromatographic
c a l i b r a t i o n d a t a c o v e r i n g a wide range w i l l show t h a t t h i s c o n d i t i o n
i s not met. Some s o r t o f treatment must be done w i t h the d a t a i n
o r d e r t o have an e q u a l - v a r i a n c e c o n d i t i o n . Transformation of
response data has been found to s a t i s f y the e q u a l v a r i a n c e
requirement ( 1 )
He c o n t i n u e s : "Often the c a l i b r a t i o n graphs a r e l i n e a r over a
p a r t i c u l a r r e g i o n but not over the r e g i o n o f i n t e r e s t . Therefore,
one s h o u l d check the l i n e a r i t i e s and use a n o n - l i n e a r c a l i b r a t i o n
graph as w e l l . " I f f i t t i n g to a l i n e a r function i s desired ( f o r
s i m p l i c i t y ' s s a k e ) , the t r a n s f o r m a t i o n f u n c t i o n used f o r t h e amount
v a r i a b l e s need n o t be the same as t h a t used on the response v a r i a b l e s
f o r e q u a l v a r i a n c e . I n t h i s manner d a t a can o f t e n be brought i n t o
c o n f o r m i t y w i t h the s i m p l e r l i n e a r f u n c t i o n C O .
"Okay, what a r e the b e t t e r s o l u t i o n s ? Use e q u a l v a r i a n c e space.
Put c o n f i d e n c e l i m i t s around t h a t graph i n e q u a l v a r i a n c e space. You
may have t o do a t the b e g i n n i n g a l i t t l e more c a l i b r a t i o n , make a few
more s t a n d a r d s , and put a l i t t l e more work i n t o e s t a b l i s h i n g those
c a l i b r a t i o n g r a p h s . Don't t r y t o extend t h e c u r v e beyond your r e g i o n
of i n t e r e s t . I t i s n o t r e a s o n a b l e t o a s s o c i a t e the same v a r i a n c e
w i t h t h e b l a n k s t h a t you a s s o c i a t e w i t h t h e c u r v e . And t h e r e f o r e ,
one would have t o do a l i t t l e more work w i t h the b l a n k s and determine
what t h e d e c i s i o n l i m i t i s b a s e d o n . I f the d i l u t i o n e r r o r i s
s i g n i f i c a n t , determine the ' d e c i s i o n l i m i t ' on the b a s i s o f b l a n k s
alone
"What w o u l d t h e i d e a l system c o n s i s t o f ? W e l l , f i r s t o f a l l we
would determine what t h e c o s t o f the Type I e r r o r ( f a l s e p o s i t i v e ) i s
and what the c o s t o f a Type I I e r r o r ( f a l s e n e g a t i v e ) i s . Then we
would determine the p r o b a b i l i t y o f e n c o u n t e r i n g those particular
errors. The t h i r d s t e p i s t o determine the d i s t r i b u t i o n o f f u t u r e
samples (whether u n i f o r m , n o r m a l , o r w h a t ) . F i n a l l y one can m i n i m i z e
the p r o b a b l e c o s t o f f u t u r e e r r o r s . "
These l a b o r a t o r y problems as r e l a t e d i n t h i s f i r s t p a r t o f the
p a n e l d i s c u s s i o n a r e i m p o r t a n t t o s o l v e i n o r d e r t o o b t a i n good
w o r k i n g d a t a . However, when the d a t a g e t s put i n t o the p u b l i c a r e n a ,

look out f o r more problems. A few are d e s c r i b e d i n the next

sections.
Limit of Detection and Public Trust (Dr. W. B. Crummett)
"Over the past several years great progress has been made by the
s c i e n t i f i c community working largely through s c i e n t i f i c societies to
define and understand the meaning of the ' l i m i t of detection' and the
' l i m i t of q u a n t i f i c a t i o n ' . Thus, the American Chemical Society has
issued guidelines (2) and principles (3) of environmental a n a l y s i s .
The A s s o c i a t i o n of O f f i c i a l A n a l y t i c a l Chemists (AOAC) and the
American Society for Testing Materials (ASTM) have continued to
emphasize collaborative studies, cooperating with the International
Union of Pure and Applied Chemistry (IUPAC) i n holding symposia and
studying best ways to conduct such studies," writes Dr. Crummett.
He continues: "In
analysts i n the r e a l worl
which sometimes cause th y questione
the meaning to be misinterpreted. The problem becomes much more
serious as the ' l i m i t of detection' i s approached." Either the
c r e d i b i l i t y of science i s put into question or the wrong impression
i s given to the public i n one manner or another.
The biggest problem with interpretation by the public i s the
general public lack of understanding about uncertainty. If I have
three apples i n my basket, then the person on the street knows there
are three apples there because he can count them. However, i f a
measurement of 2 ppm for an analysis of pesticide "x" i n my drinking
water i s reported, he expects that not only was there exactly 2 ppm
in the sample but that there i s exactly and always 2 ppm there, now
and anytime l a t e r . Our schools simply do not teach uncertainty and
change, but they should...way down i n the lower grades.
The f i r s t example Dr. Crummett talks about refers to analyses of
dioxin compounds i n human f a t . Ten years ago analyses were done at
the ppm l e v e l , and- there was uncertainty i n those figures, both i n
i d e n t i f i c a t i o n and i n q u a n t i f i c a t i o n . More recently, levels were
being reported i n the high ppb range. At this time quantities of
dioxin i n the medium and low ppt (parts per t r i l l i o n , picograms/g)
ranges are being reported. These are 1000 to 1,000,000 times more
sensitive than the best analyses previously done. Here i s the f i r s t
example :
"Look at the results i n a study sponsored by the Environmental
Protection Agency, EPA, and conducted by P h i l Albro of the National
Institute of Environmental Health Sciences, NIEHS ( 4 ) . Eight i n t e r -
nationally known laboratories participated i n the study, each using
their favorite method.
Human fat was spiked with various concentrations of three PCDD's
and three PCDF's at low parts per t r i l l i o n l e v e l s . As part of the
data set, the number of unusually low and unusually high values were
reported. These were the number of values that deviated by 50% from
the spiked amount i n the sample." A t o t a l of 54 samples were
reported i n this example from each laboratory. The data are found i n
Table I.
The v a r i a b i l i t y of analysis i n this case i s quite pronounced,
but i t may be understandable i n view of the extremely low l e v e l of

Table I . EPA/NIEHS C o l l a b o r a t i v e Study o f D i o x i n i n Human F a t .

Number o f V a l u e s Above and Below 50% o f Spiked
Amount i n 54 A n a l y s e s
A n a l y t i c a l Method, L a b o r a t o r i e s
1 2 3 4 5 6 7 8
No. of Low V a l u e s 0 0 0 1 10 8 7 0
No. i n Mid Range 50 54 52 47 42 43 31 51
No. of High Values 4 0 2 6 2 3 16 3
a n a l y t e p r e s e n t and the number o f a n a l y s i s s t e p s t h e p r o c e s s r e -

quired. Dr. Crummett c o n c l u d e s i n t h i s c a s e : " A l t h o u g h such p e r f o r -
mance i s s u f f i c i e n t f o r some s c i e n t i f i c s t u d i e s i n which t r e n d s a r e
the g o a l , i t i s not q u a n t i t a t i v h f o decision-makin whe
regulation i s indicated.
Data o f t h i s s o r t
i f the proper s t a t i s t i c a l c a l c u l a t i o n f o r the d e c i s i o n l i m i t and t h e
l i m i t o f d e t e c t i o n i s performed (_5). I n r e g u l a t o r y cases a t t e n t i o n
i s g i v e n t o the s t a n d a r d d e v i a t i o n s o f t h e b l a n k and t h e a n a l y t e
d e t e c t e d i n t h e r e g i o n o f the l i m i t . Similar risks for false posi-
t i v e s and f a l s e n e g a t i v e s ( a and 3) s h o u l d be s p e c i f i e d . As i t t u r n s
o u t , however, i n the example above c i t e d , t h e r e were p r o b a b l y l a b o r a -
t o r y procedures a f f e c t i n g t h e v a r i a b i l i t y t h a t s h o u l d have been
examined b e f o r e t h e d a t a were s u b m i t t e d f o r r e g u l a t o r y purposes.
A n a l y s i s problems f o r d i o x i n a t the p a r t s p e r t r i l l i o n l e v e l i n
human f a t would be s i m i l a r t o p a r t s per q u a d r i l l i o n (1000 times more
remote) i n water because w i t h water one works u s u a l l y w i t h a much
l a r g e r sample s i z e . S i n c e few work a t the p a r t s per q u a d r i l l i o n
l e v e l i n w a t e r , i t i s expected t h a t the d i o x i n problem s h o u l d be
extremely severe. Another s t u d y quoted by Dr. Crummett i n v o l v e d
a n a l y s i s a t t h i s l e v e l i n f i v e p r i v a t e water w e l l s performed by f o u r
d i f f e r e n t l a b o r a t o r i e s . As Dr. Crummett r e l a t e s :
"Some o f the consequences and c o n f u s i o n o f u s i n g such methodol-
ogy can be seen from a s t u d y the EPA attempted i n 1984 t o determine
i f 2,3,7,8-TCDD had contaminated p o t a b l e water i n t h e M i d l a n d , MI
a r e a . The r e s u l t s o f t h i s s t u d y a r e shown i n Table I I .
Table I I . 2,3,7,8-TCDD A n a l y s i s i n F i v e Water W e l l s performed by

Four L a b o r a t o r i e s over a P e r i o d o f Time.
A n a l y s i s i n pg/L ( l i m i t o f d e t e c t i o n , pg/L)
Sampling Date Lab 1 Lab 2 Lab 3 Lab 4

November, 1984 (a)
December, 1984 ND(5-50)
June, 1985 ND (10) ND (8) ND (10) ( a )
August, 1985 ND (1) ND ( 1 0 )
September, 1985 ND (2) ND (1) ND ( 3 )
(a) L a b o r a t o r y sample c o n t a m i n a t i o n p r e s e n t

"The f i r s t l a b o r a t o r y was r e p o r t e d to have contaminated the

f i r s t s e t of samples, and i t o b t a i n e d out of c o n t r o l d e t e c t i o n on the
second s e t . A t h i r d sampling was sent to t h r e e d i f f e r e n t l a b o r a
t o r i e s , u s i n g s u p e r i o r methodology. S t i l l , some c o n t a m i n a t i o n and
v a r i a b i l i t y was e x p e r i e n c e d , and the w e l l s had t o be sampled t w i c e
more. The f o l l o w i n g f r o n t page h e a d l i n e i n the M i d l a n d D a i l y News
o f f e r e d one r e s o l u t i o n t o e r u d i t e p e o p l e who may have s e e n t h e
v a r i a b l e l i m i t s of d e t e c t i o n f o r these a n a l y s e s : 'EPA Says those who
1
Don't T r u s t W e l l Water Should Buy i t i n B o t t l e s . (6)
" I t c a n be shown t h a t p e o p l e c a n have a g r e a t c o n c e r n when
v a r i a b l e sampling and a n a l y t i c a l work i s known t o them. However,
t h i s example a l s o shows a l a c k of u n d e r s t a n d i n g on the p a r t of the
media and the p u b l i c c o n c e r n i n g the u n c e r t a i n t y and d i f f i c u l t y of
such an a n a l y s i s . "
Dr. Crummett concludes h i s d i s c u s s i o n , "The use of i n a p p r o p r i a t e
sampling a n a l y t i c a l methodolog
numbers i n t o the d a t a s e
be communicated to the p u b l i c Misinterpretation y g
of the l a b o r a t o r y r e p o r t may a l s o cause h e a d l i n e s i n newspapers which
convey a message c o n t r a r y to the d a t a i n the i n i t i a l l a b o r a t o r y
report.
" A l l f a l s e n e g a t i v e r e s u l t s thus generated l e a d t o a f a l s e sense
of s e c u r i t y . F a l s e p o s i t i v e r e s u l t s , on the o t h e r hand, l e a d t o an
e x p e n s i v e c o u r s e of i n v e s t i g a t i o n which i s sure to i n v o l v e one or
more governmental a g e n c i e s and produce unwarranted f e a r i n the gener
a l p u b l i c . Are we a n a l y t i c a l chemists t o t a l l y r e s p o n s i b l e f o r t h i s
unnecessary a c t i v i t y and f e a r ? We are p r o b a b l y n o t , a l t h o u g h we are
o f t e n accused of i t . "
L i m i t of D e t e c t i o n R e v i s i t e d ( D r . D. A. Kurtz)
At t h i s p o i n t , perhaps some e f f o r t should be made i n s i m p l i f i e d terms

to d e s c r i b e a method of c a l c u l a t i n g a l i m i t of d e t e c t i o n i n s t a t i s
t i c a l terms f o r t y p i c a l l y generated c a l i b r a t i o n d a t a . The b a s i c
problem of t h i s l i m i t , LOD, i s t h a t i t i s not b e i n g d e a l t w i t h i n the
o v e r a l l measurement. An LOD i s a s i m p l e number c o n t a i n i n g s t a t e d
r i s k s so t h e r e i s no danger i n r e p o r t i n g i t to the p u b l i c . Its
and 3 have t o be shown, and t h i s c o u l d be done i n terms of i t s
r i s k . Too f r e q u e n t l y the d e t e c t i o n d e c i s i o n l e v e l ( " c r i t i c a l l e v e l " )
and the d e t e c t i o n l i m i t are c a l c u l a t e d to be one and the same w i t h
the r e s u l t t h a t the f a l s e n e g a t i v e ends up h a v i n g a r i s k of 50%,
infra.
Dr. C u r r i e has d e s c r i b e d t h i s i n simple terms (5): "Our b a s i c
t a s k i s to d i s t i n g u i s h the b l a n k or background from a t r u e s i g n a l at
the d e t e c t i o n l i m i t . [That] can be done, p r o v i d e d t h a t the s i g n a l s
are random, independent, and s t a t i o n a r y . To c o m p l e t e l y s p e c i f y the
f a l s e p o s i t i v e (a) and f a l s e n e g a t i v e (3) r i s k s , we must know the
form of the [ s i g n a l ] d i s t r i b u t i o n and i t s p a r a m e t e r s . For most
a n a l y t i c a l s i t u a t i o n s we assume the d i s t r i b u t i o n to be normal
( G a u s s i a n ) , and the d i s p e r s i o n parameter i s s i m p l y the i m p r e c i s i o n
(standard d e v i a t i o n ) . "
F o l l o w i n g t h i s argument we must f i r s t a c c e p t the l e v e l s of r i s k ,
and 3 C o n v e n t i o n a l l y , we choose e q u a l r i s k s a t the 5% l e v e l .
Hence, = 3= 5%. I f we s e l e c t Student's t - s t a t i s t i c s i n p l a c e of

z - s t a t i s t i c s because of l i m i t e d sample s i z e , we d e f i n e the d e c i s i o n

l e v e l ( L ) and d e t e c t i o n l i m i t ( L ~ ) as shown i n the f o l l o w i n g s k e t c h :
r
0 L c L D E s t i m a t e d Net S i g n a l
The e q u a t i o n s of the two l i m i t s are as follows:
L = (la)
C -
(lb)
The s t a n d a r d d e v i a t i o n
d e v i a t i o n o f the net sample e s t i m a t e () near the l i m i t o f d e t e c t i o n
0
by the r e l a t i o n :
- a / 1 + 1/n
where i s the number of r e p l i c a t e s used t o e s t i m a t e the b l a n k ( B ) .

S i n c e the r i s k s and b are chosen t o be of e q u a l v a l u e , the c a l c u l a
t i o n f o r L i s s i m p l y 2 L f o r the case when the s c a l e r e f e r s t o the
p
net ( b l a n k - c o r r e c t e d ) s i g n a l ( i . e . , has a mean of z e r o ) . Observed

values t h a t are l e s s than L a r e s i m p l y r e p o r t e d as b e i n g n o t
d e t e c t e d because the f a l s e p o s i t i v e r i s k has been exceeded from the
chosen amount. True v a l u e s l e s s than L a r e not d e t e c t a b l e because
the f a l s e n e g a t i v e r i s k exceeds 3
L i m i t a t i o n s on the I n t e r p r e t a t i o n of F o r e n s i c L a b o r a t o r y R e s u l t s
(Mr. C. R. M i d k i f f )
The v a l u e of the l i m i t of d e t e c t i o n i n f o r e n s i c work i s shrouded i n

d e t a i l s not u s u a l l y p r e s e n t i n o t h e r t r a c e a n a l y s i s e f f o r t s . I n t h i s
a r e a the performance o f the a n a l y s t r e l a t e s t o samples s u b m i t t e d as
l e g a l e v i d e n c e , perhaps f o r some c r i m e . The sample may o f t e n be i n
such s h o r t s u p p l y t h a t o n l y l i m i t e d a n a l y s e s can be performed on i t .
In these cases the l i m i t of d e t e c t i o n can be g r e a t l y a f f e c t e d .
Mr. M i d k i f f i n t r o d u c e s t h i s t o p i c i n the f o l l o w i n g manner:
"Recently, i n c r e a s i n g concerns have been addressed about the
r e l i a b i l i t y of r e s u l t s from f o r e n s i c l a b o r a t o r i e s . One q u e s t i o n of
p a r t i c u l a r concern which may be asked i s the e x t e n t t o which these
poor r e s u l t s are a t t r i b u t a b l e to attempts t o push d e t e c t i o n l i m i t s
beyond the l i m i t s of ' r e a s o n a b l e s c i e n t i f i c c e r t a i n t y ' , making i n t e r
p r e t a t i o n of r e s u l t s tenuous.
"The f o r e n s i c chemist i s concerned not o n l y w i t h conduct of an
a n a l y s i s and p r e p a r a t i o n of a r e p o r t , but a l s o w i t h the p r e s e n t a t i o n
of the r e s u l t s i n a c o u r t of law. C o n c l u s i o n s and o p i n i o n s of the
a n a l y s t may have economic impact on i n d i v i d u a l s and o r g a n i z a t i o n s
r a n g i n g f r o m d e n i a l o f i n i t i a l employment t o l o s s o f a c u r r e n t

p o s i t i o n or the n e c e s s i t y f o r payment of s i z e a b l e monetary damages.

For i n d i v i d u a l s accused of a c r i m e , the p o t e n t i a l e f f e c t s a r e even
more s e r i o u s , i n v o l v i n g l o s s of l i b e r t y o r l i f e i t s e l f . "
I n the courtroom such a n a l y t i c a l q u e s t i o n s as the f o l l o w i n g can
be asked of the a n a l y t i c a l chemist d u r i n g c r o s s - e x a m i n a t i o n p r o c e -
d u r e s : "Were the v a l u e s which you c l a i m t o have o b t a i n e d c l o s e to the
d e t e c t i o n l i m i t f o r the method used? What i s the d e t e c t i o n l i m i t f o r
( m a t e r i a l ) i n (sample t y p e ) u s i n g (method)? Is that d e t e c t i o n l i m i t
based upon e x a m i n a t i o n of a pure sample o r one l i k e the one w h i c h you
examined i n t h i s case? Are you f a m i l i a r w i t h the IUPAC method (7)
f o r the d e t e r m i n a t i o n of the d e t e c t i o n l i m i t ? "
The Sample i n F o r e n s i c Work. " I n the c o n t e x t of c o n v e n t i o n a l a n a l y t -

i c a l c h e m i s t r y , " Mr. M i d k i f f r e l a t e s , "a method i s designed t o o p t i -
mize the d e t e r m i n a t i o n of a p a r t i c u l a r a n a l y t e i n a d e f i n e d m a t r i x .
A l t h o u g h the c o m p o s i t i o f th matri th variatio i
n o r m a l l y w i t h i n an expecte
matrix i s reproducible
the d e t e r m i n a t i o n are c o n s i d e r e d i n the d e s i g n o f the a n a l y t i c a l
method. Once a d e t e r m i n a t i o n of d e t e c t i o n l i m i t i s made, the a n a l y s t
may be c o n f i d e n t t h a t i f the m a t e r i a l of i n t e r e s t i s p r e s e n t i n the
sample above a m i n i m a l l e v e l , i t w i l l be d e t e c t e d and measured w i t h
an a c c e p t a b l e l e v e l of p r e c i s i o n and a c c u r a c y .
"In an a n a l y t i c a l s i t u a t i o n t y p i c a l of a q u a l i t y o r p r o c e s s
c o n t r o l l a b o r a t o r y , adequate sample i s p r o v i d e d f o r a l l t e s t s t o be
conducted More sample may be o b t a i n e d i f needed. I n a d d i t i o n , the
sample m a t r i x i s u s u a l l y known and i t s e f f e c t s u n d e r s t o o d .
"By c o n t r a s t , i n the f o r e n s i c l a b o r a t o r y , the q u a n t i t y of sample
i s o f t e n l i m i t e d and a d d i t i o n a l m a t e r i a l i s u n o b t a i n a b l e . As a
r e s u l t , t h e r e i s l i t t l e o p p o r t u n i t y f o r a n a l y t i c a l methodology modi-
f i c a t i o n and m a t r i x adjustment or o p t i m i z a t i o n . F u r t h e r , the m a t r i x
i t s e l f i s v a r i a b l e as i n cases i n v o l v i n g a r s o n .
" A d d i t i o n a l l y , c u r r e n t l e g a l t r e n d s r e q u i r e t h a t not a l l the
e v i d e n c e be consumed i n the t e s t i n g p r o c e s s . The c o u r t s have, i n
r e c e n t y e a r s , taken a dim v i e w of s i t u a t i o n s i n which the e n t i r e
sample of e v i d e n t i a r y m a t e r i a l was consumed d u r i n g a n a l y s i s and
u n a v a i l a b l e f o r e x a m i n a t i o n by opposing e x p e r t s . "
Matrix Effects. S i g n a l S u p p r e s s i o n and Enhancement. Major e f f e c t s

on the LOD can be found due t o the u n a v o i d a b l e presence of the sample
m a t r i x . These e f f e c t s can take the form of s u p p r e s s i o n , enhancement,
or masking o f the a n a l y t e s i g n a l . S u p p r e s s i o n and enhancement w i l l
be d i s c u s s e d here and masking i n the next s e c t i o n . Without t a k i n g
these e f f e c t s i n t o a c c o u n t , major e r r o r s can be made i n the c o n c l u -
s i o n s drawn from the e v i d e n c e .
Mr. M i d k i f f w r i t e s : " S i g n a l s u p p r e s s i o n r e f e r s t o e i t h e r a de-
c r e a s e i n the o v e r a l l a n a l y t i c a l s i g n a l o r s u p p r e s s i o n of the s i g n a l
at s e l e c t i v e a n a l y t i c a l w a v e l e n g t h s . M a t r i x - r e l a t e d s i g n a l suppres-
s i o n i s a common problem i n m i n e r a l a n a l y s i s by atomic s p e c t r o s c o p y .
In a s t u d y u s i n g I n d u c t i v e l y Coupled Plasma, ICP, atomic e m i s s i o n , i t
was found t h a t w h i l e most a n a l y t e s e n s i t i v i t i e s were depressed by
m a t r i x e f f e c t s some up t o 30%, l i t h i u m c o u l d be e i t h e r suppressed
o r enhanced depending on s m a l l changes i n c o n d i t i o n s ( 8 ) . In a
s i m i l a r s t u d y of low-power ICP, a l l m e t a l s s t u d i e d were suppressed by

i n c r e a s i n g sodium c o n c e n t r a t i o n except a r s e n i c w h i c h was enhanced

(9).
" S e l f - a b s o r p t i o n and s a t u r a t i o n e f f e c t s a l s o l e a d t o s u p p r e s s i o n
of t h e a n a l y t i c a l s i g n a l . I n gamma o r X-ray a n a l y s i s , s e l f - a d s o r p -
t i o n i s a f u n c t i o n o f sample t h i c k n e s s . S i g n i f i c a n t s e l f - a b s o r p t i o n
i n samples a n a l y z e d by X-ray e m i s s i o n c o u l d r e s u l t i n f a i l u r e t o
d e t e c t elements p r e s e n t as t r a c e s o r i n d i c a t e t h e i r c o n c e n t r a t i o n t o
be below t h e a c t u a l l e v e l .
"For n o n - d e s t r u c t i v e gamma measurements, a mock-up o f t h e o b j e c t
b e i n g examined i s o f t e n used f o r c a l i b r a t i o n purposes. Standards o f
the elements measured a r e i n c o r p o r a t e d i n t o t h e mock-up which simu-
l a t e s t h e geometry and c h e m i c a l c o m p o s i t i o n o f t h e samples as c l o s e l y
as p o s s i b l e . "
A l t h o u g h t h e a n a l y s t may be unaware o f i t s e x i s t e n c e i n t h e
sample, s i g n a l s u p p r e s s i o n s e r v e s t o decrease system s e n s i t i v i t y and
r a i s e s t h e working d e t e c t i o limit S p e c t r a l enhancement th
o t h e r hand, i n c r e a s e s t h
detection l i m i t . "Enhancemen
absorption spectroscopy (10). A v a r i e t y o f approaches, such as
Zeeman e f f e c t c o r r e c t i o n have been proposed f o r i t s e l i m i n a t i o n ( 1 1 ) .
To a v o i d a r t i f i c i a l l y h i g h r e s u l t s , c a l i b r a t i o n s t a n d a r d s must
c o n t a i n c o n c e n t r a t i o n s o f t h e enhancing s p e c i e s e q u i v a l e n t t o those
i n t h e sample. O r d i n a r y s t a n d a r d s o l u t i o n s a r e n o t r e p r e s e n t a t i v e o f
the a n a l y t i c a l s i t u a t i o n . "
S p e c t r a l o v e r l a p i s a s p e c i a l case o f enhancement. " S p e c t r a l
o v e r l a p can cause d e t e c t i o n and measurement o f a m a t e r i a l n o t p r e s e n t
T
i n the sample (10) and g i v e f a l s e p o s i t i v e * r e s u l t s . Detection
l i m i t s must be r e c o g n i z e d as b e i n g i n s e p a r a b l y l i n k e d t o s e l e c t i v i t y
because r e l i a b l e d e t e c t i o n l i m i t s cannot be e s t a b l i s h e d u n l e s s t h e r e
i s c e r t a i n t y i n what i s b e i n g measured."
M a t r i x E f f e c t s . S i g n a l Masking. S i g n a l masking r a i s e s t h e e f f e c t i v e
d e t e c t i o n l i m i t and s e r i o u s l y c o m p l i c a t e s i n t e r p r e t a t i o n o f t h e
a n a l y t i c a l data. One cause o f s i g n a l masking i s e l e v a t i o n o f t h e
background s i g n a l l e a d i n g t o u n f a v o r a b l e s i g n a l / n o i s e r a t i o s . "An
example i s t h e e x a m i n a t i o n o f swabs c o l l e c t e d from t h e hands o f a
suspected s h o o t e r . Antimony and barium a r e r e l a t i v e l y uncommon i n
nature or manufactured products but are present i n the primer
c o m p o s i t i o n o f most t y p e s o f modern a m m u n i t i o n . D u r i n g weapon
d i s c h a r g e o r h a n d l i n g , these elements a r e d e p o s i t e d on t h e hands and
can s u b s e q u e n t l y be c o l l e c t e d w i t h a c i d - m o i s t e n e d c o t t o n swabs.
U n f o r t u n a t e l y , o t h e r m a t e r i a l s p r e s e n t on t h e hands, such a s , d i r t ,
g r e a s e , o i l o r b l o o d a r e a l s o c o l l e c t e d by t h e swabbing p r o c e s s .
" I n t h e l a b o r a t o r y , t h e swabs a r e l e a c h e d w i t h n i t r i c a c i d t o
e x t r a c t t h e barium and antimony f o r FAAS a n a l y s i s . L i g h t swabbing
r e s u l t s i n i n e f f e c t i v e e x t r a c t i o n o f these elements from t h e swabs.
The a n a l y s t w i l l o b t a i n a ' f a l s e n e g a t i v e ' r e s u l t f o r t h e presence o f
f i r e a r m s d i s c h a r g e r e s i d u e s . I f l o n g e r l e a c h i n g times o r a g i t a t i o n
of t h e swab a r e used, contaminants cause a background e l e v a t i o n t o an
e x t e n t n o t r e a d i l y noted o r c o r r e c t a b l e , even w i t h s o p h i s t i c a t e d
instrumentation.
"Masking c a n a l s o o c c u r when b l i n d l y r e l y i n g upon s o p h i s t i c a t e d
instruments. F o r example, many modern atomic a b s o r p t i o n i n s t r u m e n t s
p r o v i d e f o r d i s p l a y o f o n l y one s i g n a l a t a t i m e , e i t h e r t h e back-

ground alone o r the background c o r r e c t e d s i g n a l . The a n a l y s t r e l y i n g

on automatic background c o r r e c t i o n may be unaware o f the magnitude o f
the background s i g n a l b e i n g s u b t r a c t e d (12,13)
Background s u b t r a c t i o n , r e g a r d l e s s o f system used, does n o t
change t h e s i g n a l / n o i s e r a t i o i n the system. I t cannot, and s h o u l d
n o t , be r e l i e d upon t o c o n v e r t a sample u n f i t f o r a n a l y s i s i n t o one
from which important i n f o r m a t i o n and c o n c l u s i o n s w i l l be drawn."
Matrix V a r i a b i l i t y . L i m i t s of d e t e c t i o n of v a r i o u s analytes are

dependent n o t o n l y on t h e m a t r i x but a l s o on t h e changeable q u a l i t i e s
of the m a t r i x due t o the i n c i d e n t . T h i s l a t t e r q u a l i t y i s seen i n
a d d i t i o n o f a n a l y t e s from p y r o l y s i s and the s u b t r a c t i o n o f a n a l y t e s
from a d s o r p t i o n onto p y r o l y z e d m a t e r i a l .
Mr. M i d k i f f c o n t i n u e s , "When a sample from a suspected a r s o n i s
examined by gas chromatography, a d d i t i o n a l peaks from m a t e r i a l s
p r e s e n t a t t h e scene, a f o example p l a s t i c s i t h sampl b
o b s e r v e d . These a d d i t i o n a
normally relied upon f o
l i q u i d s i n the d e b r i s . S i m i l a r problems may be encountered i n the
a n a l y s i s o f samples from a bomb scene where c h e m i c a l s i n s o i l o r
d e b r i s from the bomb c r a t e r c o m p l i c a t e t h e d e t e c t i o n and i d e n t i f i c a -
t i o n o f e x p l o s i v e components.
"When t h e m a t e r i a l s o u g h t i s n o t d e t e c t e d i n an a d s o r b e n t -
c o l l e c t e d sample, as i n the case o f a c h a r c o a l sample produced by
f i r e , p a r t i a l o r t o t a l r e t e n t i o n by the a d s o r b e n t may be a f a c t o r .
C o n v e r s e l y , t h e adsorbent may have v a r y i n g e f f e c t i v e n e s s i n r e t a i n -
i n g t h e a n a l y t e because o f s u r f a c e i n a c t i v a t i o n , water s a t u r a t i o n , o r
some o t h e r cause (14) ."
Conclusions. Mr. M i d k i f f concludes w i t h these remarks. " I n t h e

f o r e n s i c l a b o r a t o r y , t o ensure a r e a s o n a b l e i n t e r p r e t a t i o n o f t h e
a n a l y t i c a l d a t a , t h e a n a l y s t must have c e r t a i n t y i n what i s b e i n g
measured. There must be a f o c u s on a n a l y t i c a l f u n d a m e n t a l s , and t h e
a n a l y s t must be c o g n i z a n t o f f a c t o r s i m p a c t i n g d e t e c t i o n and measure-
ment. These f a c t o r s i n c l u d e m a t r i x e f f e c t s i n t h e a c t u a l sample;
p o t e n t i a l c o n t a m i n a t i o n o r l o s s d u r i n g sample h a n d l i n g , s t o r a g e , and
a n a l y s i s ; and i n s t r u m e n t s e n s i t i v i t y . The a n a l y s t must be s u f f i -
c i e n t l y f a m i l i a r w i t h the m a t e r i a l t o determine t h e p r o b a t i v e v a l u e
of s i m i l a r i t i e s and t h e d i s c r i m i n a t i o n v a l u e o f minor d i f f e r e n c e s .
The a n a l y s t s h o u l d r e c o g n i z e t h a t p r e s e t c r i t e r i a may negate the need
f o r f i n e r d a t a and a v o i d a tendency t o attempt t o use lower and lower
values. The next d e c i m a l p l a c e may be uncharted ground and j u s t
because i t c a n be measured, does n o t g i v e i t p r o b a t i v e v a l u e . While
improved d e t e c t i o n l i m i t s o f f e r advantages i n t h e e x a m i n a t i o n of
p h y s i c a l e v i d e n c e , the a d d i t i o n a l d a t a r e q u i r e s c a r e i n i n t e r p r e t a -
tion. The r e s u l t s may have a major impact a f f e c t i n g an i n d i v i d u a l ' s
f o r t u n e , freedom, o r l i f e . To ensure f a i r n e s s and overcome q u e s t i o n s
about t h e i r c r e d i b i l i t y , f o r e n s i c s c i e n t i s t s must c o n s i d e r t h e
problems t h a t a r e c r e a t e d by improved d e t e c t i o n l i m i t s and t h e i n t e r -
p r e t a t i o n of r e s u l t s i n the f o r e n s i c l a b o r a t o r y . "

The Role o f the Blank i n t h e Measurement o f D e t e c t i o n L i m i t s i n

Atomic Spectrometry ( D r . R. L. W a t t e r s , J r . and L. J . Wood)
Some o f t h e problems i n f o r e n s i c a n a l y s i s r e l a t e d t o sample m a t r i x

and d e t e c t i o n l i m i t s can be found i n o t h e r a n a l y t i c a l l a b o r a t o r i e s as
well. B e s i d e s t h e enhancement and s u p p r e s s i o n e f f e c t s on a n a l y t e
s i g n a l s d e s c r i b e d by Mr. M i d k i f f i n the p r e v i o u s s e c t i o n , Dr. Watters
and Ms. Wood o f t h e N a t i o n a l B u r e a u o f S t a n d a r d s p r o v i d e some
s p e c i f i c i n f o r m a t i o n r e g a r d i n g m a t r i x e f f e c t s on s p e c t r a l background.
Background l e v e l and s p e c t r a l s t r u c t u r e comprise t h e c h i e f source o f
measured s i g n a l when b l a n k s a r e b e i n g measured f o r d e t e c t i o n l i m i t
estimation. T h e i r example i s taken from i n d u c t i v e l y c o u p l e d plasma
(ICP) spectrometry.
Dr. Watters w r i t e s : "A d e t e c t i o n l i m i t (DL) i s u s u a l l y used f o r
comparison, and the s p e c i f i c type o f comparison o f t e n d i c t a t e s how
the DL i s measured o r c a l c u l a t e d Whe v a l i d b l a n k measurement d
c a l c u l a t i o n s a r e made,
r e s u l t i n g DL i s a p p l i c a b l
and c o n d i t i o n s used f o r the DL d e t e r m i n a t i o n . Erroneous c o n c l u s i o n s
f r e q u e n t l y a r i s e when t h i s same DL i s used as a comparison f o r o t h e r
systems o r c o n d i t i o n s . "
One way t o ensure t h a t DL's a r e used o n l y f o r v a l i d comparisons
i s t o examine the s p e c i f i c r e a s o n f o r a p a r t i c u l a r DL e s t i m a t i o n .
Watters and Wood c o n t i n u e : "To examine t h e v a l i d i t y o f DL c o m p a r i -
sons, we can pose two q u e s t i o n s t h a t a r e supposed t o be answered by a
DL e s t i m a t e : ( 1 ) I s one i n s t r u m e n t b e t t e r than another f o r d e t e r m i n -
i n g a t r a c e a n a l y t e ? ( 2 ) Can an a n a l y t e be determined i n a s p e c i f i c
sample m a t r i x u s i n g a s p e c i f i c method?
"The u s u a l approach t o answering e i t h e r o f these q u e s t i o n s i s
based on e s t a b l i s h i n g t h e i n s t r u m e n t response w i t h r e s p e c t t o a n a l y t e
c o n c e n t r a t i o n and measuring t h e v a r i a b i l i t y o f t h i s response when no
analyte i s present. Some a n a l y s t s p r e f e r t o c a l l t h e m e a s u r e d
1
response t h e ' b l a n k . The s o l u t i o n used f o r the measurement i s a l s o
c a l l e d the 'blank' , and i t i s the c o m p o s i t i o n o f t h i s b l a n k t h a t
o f t e n l e a d s t o t h e problems mentioned above. For s i m p l i c i t y of d i s -
c u s s i o n , we w i l l c o n s i d e r samples and b l a n k s t o be s o l u t i o n s .
" A l t h o u g h the d i s t i n c t i o n i s seldom made, t h e r e a r e a t l e a s t
t h r e e d i f f e r e n t types o f ' z e r o - a n a l y t e ' samples t h a t a r e c a l l e d
'blanks'. The f i r s t i s t h e s o l u t i o n t h a t c o n t a i n s o n l y r e a g e n t s
mixed i n q u a n t i t i e s t h a t a r e used i n the c a l i b r a t i o n s t a n d a r d s o r t h e
f i n a l sample d i l u t i o n . T h i s ' s o l v e n t ' o r 'reagent' b l a n k i s some-
times used as the 'zero s t a n d a r d ' t o d e f i n e t h e y - i n t e r c e p t o f t h e
c a l i b r a t i o n curve.
"When samples a r e prepared u s i n g d i s s o l u t i o n methods, t h e t r u e
a n a l y t i c a l b l a n k c o n s i s t s o f a l l r e a g e n t s and s t e p s used i n t h e
method. The o n l y a n a l y t e p r e s e n t i n t h i s second type o f b l a n k i s
caused by c o n t a m i n a t i o n from any reagent o r c o n t a c t w i t h l a b o r a t o r y
environment and a p p a r a t u s . The l e v e l o f a n a l y t e i n t h i s ' a n a l y t i c a l '
b l a n k and i t s v a r i a b i l i t y a r e key q u a n t i t i e s t o be e v a l u a t e d i n
a c c u r a t e t r a c e a n a l y s i s ( 1 6 ) . The c o n t e n t o f the a n a l y t i c a l b l a n k i s
more method dependent than t h a t o f the reagent b l a n k .
"The t h i r d type o f s o l u t i o n b l a n k i s one t h a t c o n t a i n s e v e r y
component o f a s p e c i f i c s a m p l e e x c e p t t h e a n a l y t e o f i n t e r e s t .
R a r e l y does one encounter such a s o l u t i o n i n t h e normal c o u r s e o f

16. KURTZ ET AL. Real-World Limitations to Detection 301
analysis, and indeed i t i s d i f f i c u l t to make such a solution by

design. The usefulness of having this type of 'matrix* blank w i l l be
discussed i n terms of detection l i m i t measurements using inductively
coupled plasma (ICP) spectrometry as an example."
The example chosen to demonstrate the s i g n i f i c a n t differences i n
detection l i m i t estimates obtained by measuring d i f f e r e n t types of
blank solutions i s taken from ICP spectrometry. However, analogous
situations arise i n any form of spectrometry where net absorption or
emission peaks are measured. Dr. Waiters describes the relationship
between the questions posed above and the type of blank to be mea
sured for DL estimation:
"If one wishes to answer question 1 above, replicate ICP mea
surements are made with or without spectral background correction,
while aspirating a solution of d i l u t e pure acid i n d i s t i l l e d water.
The choice of whether or not to use background correction may be
dictated by the ICP spectromete
are not capable of backgroun
the analyte spectral peak
l i m i t require background correction. Detection l i m i t s should there
fore be estimated with net intensity measurements of the blank. In
t h i s case, the q u e s t i o n of instrument comparison i s adequately
answered by measuring the reagent blank, provided a l l measurement
parameters are r e a l i s t i c and similar from instrument to instrument.
On the other hand, i f one has to answer question 2, DL estimates
using the reagent blank can y i e l d erroneously low DL's and use of an
a n a l y t i c a l or matrix blank i s more appropriate. The reason more
complex solutions result i n higher detection l i m i t s i s that certain
components of variance are proportional to the l e v e l of signal being
measured. Any increase i n background l e v e l , spectral l i n e caused by
contamination, or spectral l i n e feature caused by an i n t e r f e r i n g
matrix element l i n e can r e s u l t i n an i n c r e a s e i n the d e t e c t i o n
limit." This effect may be offset i f the increase i n blank measure
ment variance i s compensated by matrix suppression of the c a l i b r a t i o n
curve as previously mentioned by Mr. M i d k i f f .
Ms. Wood describes how certain types of DL estimates were made
u s i n g ICP spectrometry and d i f f e r e n t kinds of blank s o l u t i o n s :
"Regardless of the mathematical expression one uses to estimate a DL,
the term common to many approaches i s the standard deviation of the
background, . This term i s estimated by r e p l i c a t e measurements of
the background as s . We have examined the e f f e c t s of blank type on
the value of s, f o r a number of spectral analysis lines using the
ICP. Where differences occur from one blank type to another, the
underlying reasons f o r the differences are examined. Five d i f f e r e n t
blanks were prepared as examples of reagent, a n a l y t i c a l , and matrix
blanks. The reagent blank consisted of 1% V/V HTCL i n d i s t i l l e d
water. An a n a l y t i c a l blank was prepared using a d i s s o l u t i o n method
for s t e e l samples. Another a n a l y t i c a l blank was prepared using a
procedure suitable f o r dissolving geological samples by lithium
metaborate fusion. Matrix blanks were prepared using 99.999% pure
iron i n one and 99.999% pure aluminum i n the other. In each case, 1
g of metal was dissolved i n 100 mL of 10% V/V 12M HC1. A sequential
ICP spectrometer was used to measure net signals for the elements and
spectral l i n e s l i s t e d i n Table I I I . Ten measurements of each of the

f i v e s o l u t i o n s were made and the s t a n d a r d d e v i a t i o n s of these r e p l i -

c a t e measurements a t each wavelength are a l s o l i s t e d i n the t a b l e .
In cases where two wavelengths are l i s t e d f o r a g i v e n element, the
s e c o n d w a v e l e n g t h has b e e n i d e n t i f i e d as one w i t h no s p e c t r a l
i n t e r f e r e n c e from i r o n ( 1 7 ) .
"The r e a g e n t b l a n k can be used as an i n d i c a t i o n of the b a s e l i n e
instrumental detection limit, when i t s s t a n d a r d deviation is
m u l t i p l i e d by the a p p r o p r i a t e constant. Data from t h i s s o l u t i o n
c o u l d be used to answer q u e s t i o n 1 above, but i n a number of cases
cannot be used to answer the second q u e s t i o n . The a n a l y t i c a l b l a n k s
can be used as i n d i c a t i o n s of method d e t e c t i o n l i m i t s , w h e r e i n wet
ash d i g e s t i o n or a l k a l i f u s i o n procedures are employed to d i s s o l v e
the sample. In the case of the f u s i o n b l a n k , t h i s a n a l y t i c a l b l a n k
a l s o s e r v e s as a s p e c i f i c type of m a t r i x b l a n k s i n c e the p r i n c i p a l
c o n s t i t u e n t s i n s o l u t i o n are l i t h i u m and boron. The c o n c e n t r a t e d
i r o n and aluminum s o l u t i o n t s o l u t i o matrice that c o n t a i
the major c o n s t i t u e n t whe
these m a t r i c e s i s of c o n c e r n
answering q u e s t i o n 2 above. I n a d d i t i o n , these b l a n k s a l s o s e r v e as
a n a l y t i c a l b l a n k s s i n c e d i s s o l u t i o n procedures were employed i n t h e i r
making. In some cases the d i f f e r e n c e s i n c a l c u l a t e d d e t e c t i o n l i m i t s
among these s o l u t i o n s i s s i g n i f i c a n t . "
Dr. Watters p o i n t s out s p e c i f i c examples from the d a t a : " S e v e r a l
differences i n standard deviation f o r r e p l i c a t e net intensity
measurements between the reagent b l a n k s o l u t i o n and the o t h e r b l a n k
s o l u t i o n s can be found i n Table I I I . In g e n e r a l , one can a s c r i b e an
i n c r e a s e i n the s t a n d a r d d e v i a t i o n f o r a n a l y t i c a l and m a t r i x b l a n k s
compared to the reagent b l a n k to three p o s s i b l e c a u s e s : These are
c o n t a m i n a t i o n from the d i s s o l u t i o n p r o c e d u r e , broadband s h i f t s i n the
s p e c t r a l background caused by a m a t r i x element, and s p e c t r a l l i n e
i n t e r f e r e n c e from a m a t r i x element. Table I I I c o n t a i n s examples of
a l l t h r e e of these and t h e i r o c c u r r e n c e i s i n d i c a t e d i n the T a b l e .
S p e c i f i c examples can be understood by i n s p e c t i o n of wavelength scans
i n the r e g i o n of the b l a n k measurement.
"The d i f f e r e n c e s i n background s t a n d a r d d e v i a t i o n f o r the Ca
measurements a t 393.366 nm p r o v i d e examples of c o n t a m i n a t i o n and
spectral line interference effects. Wavelength scans f o r the f i v e
s o l u t i o n s are p l o t t e d i n F i g u r e 1. The e f f e c t of Ca c o n t a m i n a t i o n i n
the l i t h i u m metaborate f u s i o n f l u x i s most e v i d e n t . E f f e c t s of t r a c e
amounts of Ca c o n t a m i n a t i o n i n d i s s o l v i n g the aluminum and i n the
s t e e l b l a n k s o l u t i o n s are a l s o v i s i b l e i n the f i g u r e . The wavelength
scan of the i r o n s o l u t i o n c l e a r l y shows the s p e c t r a l i n t e r f e r e n c e
from the 393.3605 nm l i n e of Fe. A l t h o u g h the most s e r i o u s degrada-
t i o n i n the Ca d e t e c t i o n l i m i t o c c u r s f o r the f u s i o n b l a n k , t h i s case
i s not l i k e l y to cause p r a c t i c a l a n a l y s i s d i f f i c u l t i e s . Fusion
t e c h n i q u e s are most o f t e n employed to d i s s o l v e g e o l o g i c a l samples i n
which Ca i s u s u a l l y a major c o n s t i t u e n t . Hence t r a c e d e t e c t i o n of Ca
i s not a c o n c e r n .
"The d e t e c t i o n of Cd at 214.438 nm i s s u b j e c t to i n t e r f e r e n c e s
from Fe and A l . I n F i g u r e 2, the s p e c t r a l l i n e i n t e r f e r e n c e of the
214.445 nm l i n e of Fe i s a p p a r e n t , as w e l l as the l a r g e s l o p i n g s h i f t
caused by the presence of the h i g h c o n c e n t r a t i o n of A l . T h i s l a t t e r
e f f e c t i s even more pronounced f o r the d e t e c t i o n of Cr a t 205.552 nm.
The scan p l o t t e d i n F i g u r e 3 shows the l a r g e continuum s h i f t caused

L i BO
\ A l 10000 ppm
- ^ ^ Y Steel
1* HN0
-~ \ /
\Fe 10000 ppm
393.366 nm
F i g u r e 1. S p e c t r a l scan of ICP e m i s s i o n i n t e n s i t y v e r s u s wave-

l e n g t h i n the r e g i o n of the 393.366 nm l i n e o f Ca f o r v a r i o u s
types of b l a n k s o l u t i o n s .
F i g u r e 2. S p e c t r a l scans showing the e f f e c t of 10000 ppm Fe and

A l on the background near the 214.438 nm l i n e o f Cd.

F i g u r e 3. The e f f e c t o f 10000 ppm A l and Fe on t h e s p e c t r a l

background o f t h e 205.552 nm l i n e o f C r .

T a b l e I I I . S t a n d a r d d e v i a t i o n o f 10 r e p l i c a t e measurements o f
t h r e e types o f b l a n k s o l u t i o n s f o r a n a l y t e element
d e t e c t i o n u s i n g ICP s p e c t r o m e t r y . Standard
d e v i a t i o n s a r e e x p r e s s e d i n u n i t s o f ng/mL
Fusion Steel
Element I(nm) 1% HN0 3 Blank Blank Fe Al
a
Ca 393 .366 0.,1 1. 0 a
0.1 0.3 b
0.6
c
Cd 214 .438 9.,6 9.,4 6.5 38 b
20
c
Cr 205 .552 21 35 19 40 b
148
Cr 427 .480 22 18 23 18 20
Cu 327 .396 4.,5 3.,3 2.3 86 b
4.0
d
Mg 279 .553 0..2 0.,2 0.1 0.9 b
1.0
Mg 280 .270 1..5 3..0 0.9 1.8 3.3
Mo 202 .030
Mo 281 .615
Ni 231 .604
c
Pb 220 .353 81 70 57 148 b
250
Si 251 .611 7,.5 13 24 a
274 b
26
Si 288 .158 13 36 18 36 13
V 311 .071 3,.2 2..1 3.2 13 b
3.7
Zn 213 .856 6 .2 5,.5 5.1 36 b
8.2
a
C o n t a m i n a t i o n from r e a g e n t s o r p r o c e d u r e s
b
Matrix spectral line interference
c
M a t r i x continuum background s h i f t
d
Unidentified spectral line interference
by A l . T h i s phenomenon has been e x p l a i n e d by L a r s o n and F a s s e l ( 1 8 ) ,

who d e s c r i b e d t h e background s h i f t from r e c o m b i n a t i o n continuum i n
the spectrum o f A l . The r e c o g n i t i o n o f t h e i m p o r t a n c e o f t h i s
continuum s h i f t l e d t o t h e g e n e r a l acceptance o f t h e need f o r back
ground c o r r e c t i o n near t h e ICP a n a l y t e e m i s s i o n l i n e . Consequently,
ICP s p e c t r o m e t e r s have t h e a b i l i t y t o measure background a d j a c e n t t o
a s p e c t r a l l i n e t o c o r r e c t f o r s y s t e m a t i c e r r o r s of q u a n t i f i c a t i o n
due t o background s h i f t s . I t i s e v i d e n t here t h a t t h e same phenome
non a l s o a f f e c t s the random e r r o r component r e l a t e d t o d e t e c t i o n .
" I t i s i n t e r e s t i n g t o note t h a t t h e secondary l i n e s f o r C r , Mg,
Mo, and S i , chosen t o a v o i d i r o n i n t e r f e r e n c e , do a l l e v i a t e o r reduce
the e f f e c t s o f Fe l i n e o v e r l a p . F o r Cr and Mo, s t a n d a r d d e v i a t i o n o f
the background f o r the secondary l i n e s i s lower f o r the Fe m a t r i x
b l a n k than f o r t h e d i l u t e a c i d b l a n k . T h i s i s p r o b a b l y caused by
matrix suppression of the c a l i b r a t i o n curve as d e s c r i b e d by Mr.
Midkiff."
Dr. Watters c o n c l u d e s : " I t i s e v i d e n t t h a t t h e term ' d e t e c t i o n
1
limit can have m u l t i p l e meanings. The type o f b l a n k t h a t i s used
f o r an e x p e r i m e n t a l e s t i m a t e o f t h e d e t e c t i o n l i m i t i s r e l a t e d t o the
s p e c i f i c type o f d e t e c t i o n l i m i t d e s i r e d . We have p r e s e n t e d examples
taken from ICP s p e c t r o m e t r i c d a t a t h a t i n d i c a t e t h e need f o r c l e a n
l a b o r a t o r y sample p r e p a r a t i o n facilities, so t h a t contamination
e f f e c t s can be reduced o r e l i m i n a t e d from any type o f d e t e c t i o n l i m i t
estimation. Even under t h e b e s t l a b o r a t o r y c o n d i t i o n s , t h e problem

of a n a l y t e d e t e c t i o n i n the presence o f complex m a t r i x elements a t

h i g h c o n c e n t r a t i o n s can be e s p e c i a l l y d i f f i c u l t . Detection l i m i t
e s t i m a t i o n r e l a t e d t o a s p e c i f i c sample treatment and sample m a t r i x
r e q u i r e s the c a r e f u l s e l e c t i o n o f the a p p r o p r i a t e b l a n k t o ensure
r e a l i s t i c estimates. So c a l l e d ' i n s t r u m e n t a l d e t e c t i o n l i m i t s ' may
be u s e f u l i n comparing one i n s t r u m e n t t o another f o r samples w i t h no
interfering species. However, r e a l i s t i c analytical detection
q u e s t i o n s must be answered by r e a l i s t i c e s t i m a t i o n procedures and
blanks."
Problems w i t h R e g u l a t o r y L i m i t S e t t i n g s ( D r . W. W. Hanneman)
[*Note: The quoted o p i n i o n s expressed here a r e those o f the a u t h o r

alone]
F e d e r a l a d m i n i s t r a t i v e a g e n c i e s r e g u l a r l y have t h e t a s k o f
s e t t i n g l i m i t s of detectio mandated b legislativ Act that
require a no-detection l e v e
While a g r e a t d e a l o f o u t s i d
the a g e n c i e s , d i f f i c u l t i e s a r i s e when an A c t c o v e r s such a broad
range o f a c t i v i t i e s as c h e m i c a l a n a l y s i s . Requirements must cover
o r g a n i c s , i n o r g a n i c s , o r g a n o m e t a l l i c s , and b i o c h e m i c a l s ; they must
handle g a s e s , s o l u t i o n s , l a r g e m o l e c u l e s and s m a l l . We have a l r e a d y
seen some examples i n the s e c t i o n above. However, even i n more
common a n a l y t i c a l a r e a s , t h e r e a r e c l a i m s t h a t problems have a r i s e n
when an agency has s e t a d e t e c t i o n l e v e l t h a t may be u n a t t a i n a b l e .
Method D e t e c t i o n L i m i t . I n EPA, where much o f t h i s work was done,

e a r l y attempts a t s e t t i n g d e t e c t i o n l i m i t s were based on the s i g n a l /
n o i s e c o n c e p t . I n 1984, came the i n i t i a l e f f o r t s t o u t i l i z e s t a t i s -
t i c a l c o n c e p t s when the Method D e t e c t i o n L i m i t (MDL) was brought i n t o
use. The next s t e p was the Maximum Contaminant L e v e l (MCL). F i n a l l y ,
the P r a c t i c a l Q u a n t i f i c a t i o n L i m i t (PQL) has been proposed. In at
l e a s t some cases these l i m i t s a r e u n a t t a i n a b l e i n the r e a l w o r l d ,
a c c o r d i n g t o D r . Hanneman, who d e s c r i b e s these problems:
"Most t o x i c o l o g i s t s b e l i e v e t h a t the dose makes the p o i s o n .
N e v e r t h e l e s s , the EPA and s t a t e r e g u l a t o r y a g e n c i e s , a b e t t e d by
a n a l y s t s and the media, have promoted the zero r i s k c o n c e p t . This
concept o f f e r s some a t t r a c t i v e advantages. I t reduces the need f o r
s c i e n t i f i c judgment, and i t s i m p l i f i e s the r e g u l a t i o n o f c a r c i n o g e n s .
The l o g i c a l consequence o f a c c e p t i n g t h i s premise i s t h a t exposure
l i m i t s must be s e t t o z e r o .
"The zero r i s k concept produced the g r e a t e s t growth a r e a f o r
a n a l y s t s i n the h i s t o r y o f mankind. New methods had t o be d e v e l o p e d ,
l i m i t s o f d e t e c t i o n had t o be lowered, and compliance had t o be
demonstrated. A l l t h i s was f u e l e d by the l i n e a r e x t r a p o l a t i o n t o
zero.
" L e t ' s examine the case o f benzo(a)pyrene ( B a P ) . We c o u l d do
the same w i t h PCB's o r almost any suspected c a r c i n o g e n . As promul-
g a t e d i n 1979, EPA M e t h o d 610 (J_9) i s a p p l i c a b l e t o i n d u s t r i a l
discharges. I t was designed t o meet m o n i t o r i n g r e q u i r e m e n t s o f the
N a t i o n a l P o l l u t i o n D i s c h a r g e E l i m i n a t i o n System (NPDES) p e r m i t s . I t s
l i m i t o f d e t e c t i o n f o r BaP was s t a t e d t o be 40 p a r t s p e r t r i l l i o n
u s i n g f l u o r e s c e n c e d e t e c t i o n ( 19) One o f o u r p l a n t s had a NPDES

permit which allowed a discharge of 50 parts per t r i l l i o n of BaP. We

had to regularly demonstrate compliance.
"The stated Detection Limit of Method 610 was determined from
the signal/noise r a t i o measured i n reagent grade water. Since actual
detection l i m i t s of methods depend upon the l e v e l of interferences
rather than instrumental l i m i t a t i o n s , there i s no compelling reason
to expect that we w i l l ever be able to demonstrate compliance u n t i l
our plant discharges only reagent grade water.
1
"In 1984 the EPA ' o f f i c i a l l y recognized that one could not make
1
measurements at 'precisely the detection l i m i t so they introduced
the concept of MDL (20).
'Definition: The Method Detection Limit (MDL) i s defined
as the minimum concentration of a substance that can be
measured and reported with a 99% confidence that the
analyte concentrations are greater than zero.'"
EPA recognized that th MDL w i l l dependin instrumental
s e n s i t i v i t i e s and matri
representative MDL's determine
value l i s t e d was 23 ppt (21 ) i n the HPLC method. The reference went
on to say that "similar results were obtained using representative
waste waters. The MDL actually achieved i n a given analysis w i l l
vary depending on instrument s e n s i t i v i t y and matrix effects."
Paragraph 15.3 (20) said: "This method was tested by 16 laboratories
using reagent water, drinking water, surface water, and three indus-
t r i a l wastewaters spiked at s i x concentrations over a range of 0.1 to
425 yg/L. Single operator precision, o v e r a l l precision, and method
accuracy were found to be d i r e c t l y related to concentration of the
parameter and e s s e n t i a l l y independent of the sample matrix." Reagent
water was defined as water having no background peak i n the v i c i n i t y
of the analyte i n question. Quality control acceptance data f o r BaP
(21) at a 'detection l i m i t ' test concentration of 10 yg/L gave a
standard deviation of 4 yg/L (n=4). This amounts to a Relative
Standard Deviation (RSD) of 40%.
[Note: One of the biggest problems, however, with the MDL i s
that i t ignores errors of the second type, accepting false negatives.
The method puts a 50% confidence l e v e l on this type of error since
the mean of the d i s t r i b u t i o n of positive values i s three standard
deviations (to give 99% confidence) up from zero, the mean of the
blank. In order to properly assess values at the 99 % l e v e l i t would
be necessary to locate the MDL another 3 standard deviations away
from zero. In this position there i s the 99% confidence that when
the analyte i s there, i t i s reported and when not there, i t i s not
reported.]
Dr. Hanneman i l l u s t r a t e s the r e a l data, obtained for setting the
MDL f o r BaP, i n Figure 4 (22). "The histograms represent the raw
data from Method 610 v a l i d a t i o n , at 200, 2000, and 12,000 parts per
trillion. For BaP the RSD ranged from 40 to 53%. Even so, i n order
to bring i t down to this l e v e l , the s t a t i s t i c i a n s had to discard
about 1/5 of the data. The results at a l l levels were the same. The
method d i d not define a l i m i t of detection."
And he continues, "This prompts the question, 'What are the
performance c r i t e r i a for a v a l i d method?' The answer, i n essence, i s
...a validated method i s one that has undergone the v a l i d a t i o n proce-
dure. There i s no requirement that i t produce meaningful numbers...

and i n the case of b e n z o ( a ) p y r e n e , the method i s e s s e n t i a l l y worth-

l e s s f o r measuring a t these l e v e l s . "
Maximum Contaminant L e v e l . " I n November of l a s t year new measures

were proposed. EPA e s t a b l i s h e d the Recommended Maximum Contaminant
L e v e l , RMCL (23) , f o r c a r c i n o g e n s and s u s p e c t e d c a r c i n o g e n s . " The
RMCL i s the maximum l e v e l of a contaminant i n d r i n k i n g water a t which
no known o r a n t i c i p a t e d adverse e f f e c t on the h e a l t h of persons would
1
o c c u r and which i n c l u d e s an adequate margin of s a f e t y . RMCL s a r e
T
non-enforceable. Maximum Contaminant L e v e l s , M C L s , are e n f o r c e a b l e
s t a n d a r d s . They a r e s e t as c l o s e as f e a s i b l e t o RMCL's, w i t h the use
of the b e s t t e c h n o l o g y , treatment t e c h n i q u e s , and o t h e r means, which
the a d m i n i s t r a t o r f i n d s are g e n e r a l l y a v a i l a b l e ( t a k i n g c o s t s i n t o
consideration)
D r . Hanneman c o n t i n u e s , "While these MCL's are p r e s e n t l y t o u t e d
as d r i n k i n g w a t e r s t a n d a r d s histor teache tha permi
w r i t e r s w i l l invoke them
the atmosphere, and urba
s u r f a c e , BaP w i l l always be p r e s e n t . We are not p r e s e n t l y i n com-
p l i a n c e and can never be i n compliance."
Practical Quantification Limit. "As of now the LOD i s a dead i s s u e

as f a r as suspected c a r c i n o g e n s are concerned. A l l the a n a l y t i c a l
t e s t s have been reduced to n o t h i n g more than q u a l i t a t i v e i n d i c a t o r s .
What i s even more f r i g h t e n i n g i s t h a t t h e r e i s no way to c o u n t e r a c t a
1
single 'false positive result.
"EPA i s now p r o p o s i n g the P r a c t i c a l Q u a n t i f i c a t i o n L i m i t , PQL
(24). The PQL i s to be s e t somewhere between 5 and 10 times the
MDL."
[Note: The PQL was s e t f o r v o l a t i l e o r g a n i c compounds, VOC, i n
the f o l l o w i n g manner: Seven of the most e x p e r i e n c e d EPA and c o n t r a c t
l a b s r e p o r t e d MDL*s i n the range of 0.2 to 0.5 y g / L . When the same
t e s t was repeated i n the same or h i g h e r ranges of c o n c e n t r a t i o n
amongst 30-40 o t h e r laboratories, the number of laboratories
r e p o r t i n g c o n c e n t r a t i o n s d i f f e r e n t from the t r u e v a l u e was n o t e d .
There was a f a i r l y l a r g e number of l a b o r a t o r i e s r e p o r t i n g f i g u r e s
d i f f e r e n t from the t r u e v a l u e by more than 20%, but no more than 5
l a b s (6 i n one case) r e p o r t e d a n a l y s i s f i g u r e s d i f f e r e n t from the
t r u e v a l u e by more than 40%. Hence, i t was concluded t h a t a h i g h
percentage of l a b o r a t o r i e s c o u l d determine these compounds w i t h i n 40%
of the t r u e v a l u e . The PQL was determined by t a k i n g the h i g h end of
the 7-lab range, 0.5 y g / L , adding 40% of t h a t to o b t a i n 0.7 yg/L and
t h e n c h o o s i n g t h e m i d p o i n t o f 5 t o 10 t i m e s t h a t f i g u r e . This
amounted to 5 yg/L. Thus, t h i s f i g u r e was determined i n t h i s case
w i t h o u t s t a n d a r d d e v i a t i o n s and no RSD can be c a l c u l a t e d . ]
" I want to c l o s e by r e i t e r a t i n g what Dr. H o r w i t z demonstrated
f i v e y e a r s ago. F i g u r e 5 (25) shows the r e l a t i o n s h i p between the
i n t e r l a b o r a t o r y r e l a t i v e s t a n d a r d d e v i a t i o n s , RSD ( o r c o e f f i c i e n t s of
v a r i a t i o n ) w i t h c o n c e n t r a t i o n . The i n t e r s e c t i o n of the heavy curve
w i t h the 34% RSD l i n e i n d i c a t e s t h a t t h e r e i s no reason to b e l i e v e
t h a t one can even d e f i n e a t r u e MDL below approx. 8 ppb, l e t a l o n e at
50 p p t . The i n t e r s e c t i o n w i t h the 20% RSD l i n e i n d i c a t e s t h a t the
PQL p r o b a b l y cannot be s e t below 250 ppb. T h i s i s not 5-10 times
above the MDL, but r a t h e r 12,000 times t h a t of the 17 ppt v a l u e

TRUE CONC.
TRUE CONC
J200 PPT 12000 PPT
AVG. BIAS = -36%
F i g u r e 4. Histograms o f raw d a t a from Method 610 V a l i d a t i o n

Study.
HORWITZ C U R V E CV%=2 ( 1
" - 0 5 L O G C )
PCB
PAH

.^70%

_ D
J P ^
I i
f ^ a[ppB
34%
250 PPB
1 1
100
PPB
F i g u r e 5. H o r w i t z Curve and a n a l y t i c a l results.

decreed f o r the n o n - c a r c i n o g e n i c p o l y n u c l e a r a r o m a t i c compound, PNA,

benzo(k)fluoranthene.
" F i n a l l y , the d a t a p o i n t s shown a r e t y p i c a l of those g e n e r a t e d
1
i n the v a l i d a t i o n s t u d i e s f o r PNA s and PCB's. Mind you, these
r e s u l t s were o b t a i n e d a f t e r e l i m i n a t i n g about 20% of the r e s u l t s .
" T h e r e i s l i t t l e e v i d e n c e t h a t an RSD o f 20% c a n e v e n be
reached. However, the g r e a t e s t p i t y of a l l i s t h a t t h i s a c t i v i t y
w i l l i n no way c o n t r i b u t e t o the b e t t e r h e a l t h of the p u b l i c . "
Remarks on D e t e c t i o n L i m i t s ( D r . W. Horwitz)
When a l l i s s a i d and done about the i n t r i c a t e d e t a i l s of c o r r e c t l y

c a l c u l a t i n g l i m i t s of d e t e c t i o n and the l i k e , we need some a d v i c e t o
be a b l e to s t a n d back from what we a r e d o i n g j u s t to see the t o t a l
effect. T h i s s e c t i o n p r o v i d e s s u c h a d v i c e and i s g i v e n by D r .
H o r w i t z , l o n g a mentor i th f i e l d On f th interestin d
u s e f u l items of a d v i c e i
and to respond t o r e q u e s t
s e t t i n g p r o c e s s . The quotes below are a l l from Dr. H o r w i t z .
F i r s t , A n a l y s i s Purpose. "The f i r s t t h i n g t h a t s h o u l d always be

known i n p e r f o r m i n g any a n a l y s i s i s i t s purpose. I n almost a l l cases
when d e a l i n g w i t h a l i m i t of d e t e c t i o n o r l i m i t of d e t e r m i n a t i o n , the
p r i m a r y purpose of d e t e r m i n i n g t h a t l i m i t i s t o s t a y away from i t .
Almost by d e f i n i t i o n , the r e l i a b i l i t y of r e s u l t s near the l i m i t s of
d e t e c t i o n or d e t e r m i n a t i o n i s e x t r e m e l y poor and i n f a c t u l t i m a t e l y
r e a c h e s a p o i n t where a d e t e r m i n a t i o n o f w h e t h e r an a n a l y t e i s
p r e s e n t o r absent can be o b t a i n e d j u s t as r e l i a b l y by t o s s i n g a c o i n
as by making the so c a l l e d measurement."
Second, I d e n t i f y b e f o r e Measure. "The second i m p o r t a n t p o i n t i s a

m a t t e r of d e f i n i t i o n . There are so many p o s s i b l e d e f i n i t i o n s of
l i m i t s of d e t e c t i o n t h a t any statement u s i n g these terms s h o u l d
always be accompanied by a d e f i n i t i o n as to how the u s e r means i t .
' L i m i t of d e t e c t i o n ' can have t h r e e d i f f e r e n t p r i m a r y meanings
whether or not an a n a l y t e (1) was i d e n t i f i e d , (2) measured, o r (3)
i d e n t i f i e d and measured. A l t h o u g h measurements can be made w i t h o u t
knowing what i s b e i n g measured, they cannot be i n t e r p r e t e d w i t h o u t
t h i s i n f o r m a t i o n . The p r o p e r t y of knowing what i s b e i n g measured i s
related to the a t t r i b u t e of method s p e c i f i c i t y . A complete
a n a l y t i c a l system of s a m p l i n g , d e t e r m i n a t i o n , and interpretation
r e q u i r e s t h a t we p l a c e i d e n t i f i c a t i o n b e f o r e measurement. O f t e n i t
t a k e s c o n s i d e r a b l y more m a t e r i a l t o i d e n t i f y the presence of an
a n a l y t e than t o measure i t s amount. T h e r e f o r e , i n s p e a k i n g of l i m i t s
of d e t e c t i o n we are r e a l l y measuring o n l y a naked s i g n a l whose o r i g i n
i s obscure. O f t e n we merely assume we know what we are measuring a t
these low l e v e l s . This o b v i o u s l y creates considerable ambiguity i n
our work."
T h i r d , Read the F e d e r a l R e g i s t e r . "There has been c o n s i d e r a b l e

c r i t i c i s m w i t h r e g a r d to how r e g u l a t o r y a g e n c i e s u t i l i z e the concept
of ' l i m i t s of d e t e c t i o n ' i n v a r i o u s a p p l i c a t i o n s . As background f o r
t h i s p o i n t , I would l i k e t o mention t h a t a n a l y t i c a l chemists s h o u l d
become f a m i l i a r w i t h t h a t eminent s c i e n t i f i c j o u r n a l c a l l e d the

Federal Register. There i s considerable technical material i n the

Federal Register. Its contents include numerous methods of analysis
and j u s t i f i c a t i o n f o r these methods. Extensive reviews of the
toxicology of various chemical materials appear here, often because
the bulk i s too large to publish anywhere e l s e . These reviews have
taken many man years for compilation. They are well worth noting.
The usual procedure when an agency wishes to promulgate a regulation
i s to f i r s t publish a proposal i n the Federal Register accompanied by
an i n v i t a t i o n to comment." It should be noted that announcement of
these i n v i t a t i o n s i s printed regularly i n Chemical and Engineering
News. "Ordinarily 60-90 days are given for receipt of comments, but
this can be extended by a formal request. Even so, agencies w i l l
c o n s i d e r comments that come i n a f t e r the d e a d l i n e . Regulatory
agencies are influenced by these comments. For example, almost 10
years ago the Food and Drug Administration published a proposal known
1
as the ' SOM Document dealing with trace residues of carcinogenic
compounds i n tissues o
primarily with r i s k analysis
genesis, there was considerable a n a l y t i c a l chemistry involved. As a
r e s u l t of e x t e n s i v e comments by the a n a l y t i c a l community, that
document was recently republished as a proposal with much of the
objectionable a n a l y t i c a l material removed. Consequently, i t i s the
obligation of the professional a n a l y t i c a l chemistry community to
bring their views to the attention of the responsible regulatory
agency to avoid the promulgations of regulations not based on sound
science. Sometimes revisions are not possible, because of the way
the law i s written. An example i s the case of the Delaney Clause
which mandates a zero value. At the time i t was enacted, i t was not
realized that attaining zero was impossible. Therefore agencies are
attempting to do the best they can to adapt to the situation by
developing p r a c t i c a l zeros.
"In t h i s connection you should be i n t e r e s t e d i n a recent
proposal made by EPA for setting a p r a c t i c a l q u a n t i f i c a t i o n l i m i t for
fluoride i n drinking water. They recommend a regulatory l i m i t of 4
ppm of fluoride and a p r a c t i c a l q u a n t i f i c a t i o n l i m i t of 0.5 ppm. For
those who are familiar with fluoride determination, the quarrel with
this particular value would be that i t i s undoubtedly set too high.
You would think that p r a c t i c a l l y a l l laboratories could comfortably
determine 0.5 ppm f l u o r i d e . Yet this value was set from actual
interlaboratory studies that showed that there were s t i l l a c e r t a i n
number of laboratories that could not achieve even such a r e l a t i v e l y
l i b e r a l l i m i t of d e t e r m i n a t i o n . I suggest that EPA i s to be
congratulated i n coming up with a value for a q u a n t i f i c a t i o n l i m i t
that can be p r a c t i c a l l y achieved by most laboratories involved i n
this particular determination and recognizing that there i s no need
to go any lower (26) This concept of a p r a c t i c a l q u a n t i f i c a t i o n was
f i r s t proposed (24) i n conjunction with v o l a t i l e organic compounds
(VOC). It should be examined by a l l who are interested i n the ' l i m i t
of detection' concept."
Reporting Results. "Another important related matter i s that of

reporting low l e v e l r e s u l t s . Again we start with the purpose of the
analysis. In some cases an unambiguous answer of yes or no i s what
i s desired; i n other cases, a good approximation to the average i s

needed. The Food and Drug A d m i n i s t r a t i o n ' s t o t a l d i e t program i s a

s p e c i f i c example of the need t o know the average i n t a k e of t r a c e
n u t r i e n t s and t o x i c contaminants i n the American d i e t . A specific
number i s r e q u i r e d f o r i n t e r p r e t a t i o n by t o x i c o l o g i s t s w i t h r e s p e c t
to s a f e t y as w e l l as how t h a t v a l u e changes w i t h t i m e . Statements of
'not d e t e c t e d ' or below a c e r t a i n l i m i t of d e t e r m i n a t i o n v a l u e a r e
u n s a t i s f a c t o r y because they cannot be handled by d e s c r i p t i v e s t a t i s -
t i c s i n cases where the amounts p r e s e n t may be below any r e a s o n a b l e
d e f i n i t i o n of ' l i m i t of d e t e r m i n a t i o n , ' e s p e c i a l l y when a b l a n k
s u b t r a c t i o n i s i n v o l v e d . I n such c a s e s , we can take advantage of the
f a c t t h a t i n many i n s t a n c e s the average of a l a r g e number of v e r y
poor measurements g i v e s a remarkably good average." T h i s , of c o u r s e ,
assumes t h a t the e r r o r s are random and not b i a s e d . "That i s i n these
cases we a r e d e a l i n g p r i m a r i l y w i t h random e r r o r s of r e l a t i v e l y l a r g e
magnitude. In the l o n g run random e r r o r s c a n c e l each o t h e r out
l e a v i n g the core of the b a s i c v a l u e we are l o o k i n g f o r
"Although s c i e n t i s t
v a l u e s , these are a d i r e c
d i s t r i b u t i o n s (Figure 6). C o n s i d e r a contaminant p r e s e n t i n a food
m a t r i x a t 0.1 ppm w i t h a r e l a t i v e s t a n d a r d d e v i a t i o n (RSD) of 100%.
An RSD of 100% means t h a t one s t a n d a r d d e v i a t i o n s u b t r a c t e d from the
average g i v e s zero and any v a l u e l e s s than t h i s , say two s t a n d a r d
d e v i a t i o n s below the average, i s i n the n e g a t i v e r e g i o n . In f a c t , i n
t h i s example 16% of the measurements a u t o m a t i c a l l y w i l l be n e g a t i v e .
P h y s i c a l l y t h i s means that the blank measurement was greater than the
a n a l y t e measurement. T h i s i s a n a t u r a l consequence of u s i n g the
normal d i s t r i b u t i o n . In f a c t i f you do not get about 16% n e g a t i v e
v a l u e s i n t h i s s i t u a t i o n , t h e r e i s p r o b a b l y something wrong w i t h your
measurements. A sound o b j e c t i o n may be r a i s e d on the grounds t h a t
m e a s u r e m e n t s i n t h i s a r e a a r e n o t n o r m a l and t h i s i s a good
possibility. However, i n almost a l l cases we never have enough d a t a
to determine whether our measurements a r e normal, l o g normal, o r
f o l l o w any o t h e r type of d i s t r i b u t i o n . Because of the w e a l t h of
i n f o r m a t i o n known about the normal d i s t r i b u t i o n and the many
i n s t a n c e s where measurement d a t a have been shown to be n o r m a l l y
d i s t r i b u t e d , i t i s g e n e r a l l y best to assume n o r m a l i t y and make any
n e c e s s a r y m o d i f i c a t i o n s to the i n t e r p r e t a t i o n i f t h e r e i s evidence to
i n d i c a t e non-normality."
Summary. "To summarize the major p o i n t s : (1) The u s u a l reason we

want t o measure a d e t e c t i o n l i m i t i s t o s t a y away from i t . (2)
Report low l e v e l measurements as they come - p o s i t i v e , n e g a t i v e , or
zero - s i n c e t h e i r average i s u s u a l l y a r e a s o n a b l y good a p p r o x i m a t i o n
of a c t u a l i t y . (3) As we go down i n our c o n c e n t r a t i o n measurements,
p r e c i s i o n degenerates i n t o a q u e s t i o n of f a l s e p o s i t i v e s and f a l s e
n e g a t i v e s , and a t low enough l e v e l s a t o s s of a c o i n w i l l g i v e us
j u s t as good an answer as w i l l our most e l a b o r a t e i n s t r u m e n t a t i o n .
(4) At these low l e v e l s , you p r o b a b l y w i l l never have enough d a t a to
determine i f you have a normal or non-normal d i s t r i b u t i o n , and when
your random e r r o r i s so g r e a t , the type o f d i s t r i b u t i o n p r o b a b l y does
not make any d i f f e r e n c e anyway."

F i g u r e 6. Normal d i s t r i b u t i o n near zero showing the n e c e s s i t y t o

report negative values.

Epilogue
R e a l w o r l d problems r e s u l t i n g from a n a l y s i s of s m a l l samples a t the

l o w e s t l e v e l s are found i n f o u r a r e n a s : p e r c e p t i o n by the r e a l w o r l d
people who are not a n a l y t i c a l e x p e r t s , governmental r e g u l a t i o n s ,
i n f l u e n c e s of the c o u r t s , and p r o d u c t i o n of the a n a l y t i c a l d a t a by
the e x p e r t a n a l y s t . Each of these arenas has i t s own c h a r a c t e r i s -
t i c s , i t s own o p e r a t i n g r u l e s , and i t s own purposes. We must f i n d ,
however, enough of a common base so t h a t we can a l l l i v e a m i c a b l y i n
t h i s one w o r l d .
The person on the s t r e e t , the n o n - e x p e r t , wants a s i m p l e p i e c e
of i n f o r m a t i o n . He wants to b e l i e v e what e x p e r t s s a y . Problems i n
the p a s t , however, put doubts on what e x p e r t s have s a i d . These
doubts a r e c l o u d e d f u r t h e r by a l a c k of knowing about p r o b a b i l i t y
r i s k s and r e l e v a n c e . Dr. Crummett has spoken t o t h i s p o i n t .
The government r e g u l a t o r ( l e g i s l a t o r well administra
t o r s ) l i k e to have neat
the problem a r e a s . Th
c o m p l e t e l y d i f f e r e n t a n a l y t i c a l areas - from r a d i o a c t i v i t y c o u n t i n g ,
to chromatography, and to e m i s s i o n and a d s o r p t i o n a n a l y s i s - t h a t
t h i s approach o f t e n becomes i n e f f e c t i v e . Dr. Hanneman's e x p e r i e n c e
suggests t h a t r e g u l a t i o n s s h o u l d be problem s p e c i f i c and p h y s i c a l l y
t e s t e d under r e a l c o n d i t i o n s b e f o r e p u t t i n g them i n t o use. Rules
s h o u l d , p e r h a p s , be b r o a d and u n d e r s t a t e d . They w o u l d be made
s p e c i f i c o n l y where unique problems are i d e n t i f i e d t h a t r e q u i r e
attachment of s p e c i f i c n u m e r i c a l r e q u i r e m e n t s .
The c o u r t s have a v e r y d i f f i c u l t time h a n d l i n g a n a l y t i c a l i n f o r -
m a t i o n , as Mr. M i d k i f f has e x p l a i n e d . The c o u r t s l o o k a t i n f o r m a t i o n
i n a p r e c i s e way, b u t t h e y d o n ' t l i k e e r r o r . E v e r y t h i n g has a
s p e c i f i c cause and i t j u s t can't be " p o s s i b l y " t h e r e . A s i d e from
e d u c a t i n g a l l the judges and a t t o r n e y s w i t h s t a t i s t i c a l t e c h n i q u e s ,
the b e s t e f f o r t i s to a s s u r e t h a t the e x p e r t t e s t i m o n y p r o v i d e s v e r y
c a r e f u l l y determined i n f o r m a t i o n .
Most of the i n f o r m a t i o n p r e s e n t e d i n t h i s p a n e l touches on the
r o l e of the a n a l y t i c a l t e c h n i c i a n and r e s e a r c h e r . These i n d i v i d u a l s
need to be more s e n s i t i v e to the v a r i o u s amounts of e r r o r t h a t a r e
p r e s e n t i n each p a r t of each a n a l y s i s . I t goes a l o n g w i t h the common
q u i p i n d a t a h a n d l i n g , "garbage i n , garbage out." I f the a n a l y t i c a l
d a t a coming from the l a b o r a t o r y i s p o o r l y c o n t r o l l e d or l a c k s the
e s t i m a t i o n s of e r r o r , i t c a n ' t be used by o t h e r s f o r i n f o r m a t i o n
l e a d i n g to m e a n i n g f u l d e c i s i o n s and a c t i o n . The f i r s t s u g g e s t i o n ,
coming from Dr. T a y l o r i s to always put c o n t r o l c h a r t s i n t o each
separate t e s t . T h i s t e l l s how the p r o g r e s s of the a n a l y t i c a l t e s t s
i s going. Each p e r s o n , each t e s t , and each use of a t e s t needs the
control chart. Those u s i n g c a l i b r a t i o n g r a p h s , as Dr. S t u r d i v a n
recommends, s h o u l d handle the i n f o r m a t i o n so t h a t t h e r e i s c o n s t a n t
v a r i a n c e a t each l e v e l a c r o s s the g r a p h .
D r . Crummett has suggested t h a t the a n a l y s t must c l e a r l y s t a t e
t h e r i s k s o f m a k i n g a f a l s e d e t e r m i n a t i o n when t r u e and a t r u e
d e t e r m i n a t i o n when f a l s e . This i s e s p e c i a l l y important i n s t a t e - o f -
the-art determinations. Researchers need t o be t e r r i b l y c a r e f u l o f
the b l a n k and i t s accompanying u n c e r t a i n t i e s , as p o i n t e d out by Dr.
Watters. The r i s k s of making b o t h the f a l s e p o s i t i v e and the f a l s e
n e g a t i v e s h o u l d be c l e a r l y d e f i n e d and when so done, s i g n a l s found t o

be less than the decision l i m i t ( c r i t i c a l l e v e l ) should simply not be

reported as present. Blank signals should also be reported.
Forensic work, as implied by Mr. Midkiff, requires a great deal of
experience because of the wide sample types. The matrix i n such work
also has a profound effect on the signal - i t may enhance and i t may
mask the signal - and i s variable as well. F i n a l l y , Dr. Horwitz
suggests: 1. know the purpose for an analysis before you measure and
2. be aware of what the government has said about the topic.
On the other hand, error calculations must be s i m p l i f i e d . There
i s a u t i l i t y i n using s i m p l i f i e d , p r a c t i c a l approximations, but i t
must be certain that such approximations are clearly noted when used.
A restatement of the key questions as key actions i s appro
priate :
*** Put a quality label on data
*** Teach the public the concept of error
*** Publish data, takin int account d let
the chips f a l l a
*** Be extremely carefu
that unequivocal actions can be taken on account of
them
*** Interact with federal regulators i n the making of
suitable regulations.
The basic goal i n trace analysis work i s to be able to f i r s t
report the numbers. The second goal i s to indicate the error of each
number. The third goal i s for the user of the information, whether
i t be an agency, a newspaper reader, or the l i k e , to accept the
reports as having error, look for i t , and then use those facts to
give information and to form knowledge (27).
Acknowledgment
This paper i s published as Journal Series Paper No. 7609 of the

Pennsylvania A g r i c u l t u r a l Experiment Station.
Literature Cited
1. Kurtz, D. .; Rosenberger, J. L.; Tamayo, G. J. In "Trace

Residue Analysis"; D. A. Kurtz, Ed.; ACS Symposium Series No.
284, American Chemical Society: Washington, D. C., 1985; pp.
133-166.
2. Keith, L. H.; Crummett, W. B.; Deegan, J., Jr.; Libby, R. .;
Taylor, J. K.; Wentler, G. Anal. Chem. 1983, 55, 2210.
3. MacDougall, D.; Crummett, W. B. Anal. Chem. 1980, 52, 2242.
4. Albro, P. W.; Crummett, W. B.; Dupuy, A. E., Jr.; Gross, M. L.;
Hanson, M.; Harless, R. L.; Hileman, F. D.; Hilker, D.; Jason,
C.; Johnson, J. L.; Lamparski, L. L.; Lau, . P-Y.; McDaniel, D.
D.; Mechan, J. L.; Nestrick, T. J.; Nygren, M.; O'Keefe, P.;
Peters, T. L.; Rappe, C.; Ryan, J. J.; Smith, L. M.; Stalling,
D. L.; Weerasinghe, N. C. .; Wendling, J. M. Anal. Chem. 1985,
57, 2717-2725.
5. Currie, L. A. In "Trace Residue Analysis, Chemometric
Estimations of Sampling, Amount, and Error"; D. A. Kurtz, Ed.;
ACS Symposium Series No. 284, American Chemical Society:
Washington, D. C., 1985; pp. 50-51.

6. Gray, K. "EPA Says Those Who Don't Trust Well Water Should Buy
It i n Bottles," Midland Daily News, December 20, 1985: Midland,
MI.
7. Long, G. L.; Winefordner, J. D. Anal. Chem. 1983, 55,
712A-724A.
8. Thompson, M.; Ramsey, M. H. Analyst 1985, 110, 1413-1422.
9. Urh, J. J. Amer. Lab. 1986, 18, 105-113.
10. Lovett, R. J.; Welch, D. L.; Parsons, M. L. Appl. Spectros.
1975, 29, 470-477.
11. Koizumi, H. Anal. Chem. 1978, 50, 1101-1105.
12. Herber, R. F. M.; DeBoer, J. L. M. Anal. Chim. Acta 1979, 109,
177-179.
13. Siemer, D. D. Anal. Chim. Acta 1980, 119, 379-382.
14. Reeve, V.; Jeffery, J.; Weihs, D.; Jennings, W. J., Forensic
Sci. 1986, 31, 479-488.
15. Currie, L. ., Anal. Chem. 1968, 40, 586-593.
16. Murphy, T. J., In
Handling, Analysis
509-539.
17. Snyder, S. C., ASTM Standardization News, April 1985, 35.
18. Larson, G. F.; Fassel, V. ., Appl. Spectros. 1979, 33, 592-599.
19. Federal Register 1979 (Dec. 3), 44, 69514-7.
20. Federal Register 1984 (Oct. 26), 49, 43348.
21. Federal Register 1984 (Oct. 26), 49, 43349.
22. "EPA Method Study 20, Method 610-PNA's, Polynuclear Aromatic
Hydrocarbons", Environmental Protection Agency, 1984.
23. Federal Register 1985 (Nov. 13), 50, 46880-900.
24. Federal Register 1985 (Nov. 13), 50, 46906-7.
25. Horwitz, W. Anal. Chem. 1982, 54, 67A-76A.
26. Federal Register 1986 (April 2), 51, 11396-11412.
27. Zervos, C. In "Trace Residue Analysis"; D. A. Kurtz, ED.; ACS
Symposium Series No. 284, American Chemical Society: Washington,
D.C., 1985; pp. 235-252.

Chapter 17
Reporting Low-Level Data

for Computerized Data Bases
1 2 1 3
Martin W. Brossman , Gerard McKenna , Henry Kahn , Donald King ,
4 5
Robert Kleopfer , and John K. Taylor
1
U.S. Environmental Protection Agency, Washington, DC 20460
2
U.S. Environmental Protection Agency, Edison, NJ 08837
3
Ministry of the Environment, P.O. Box 213, Rexsdale,
Ontario M9W 5L1, Canada
4
U.S. Environmental Protection Agency, Kansas City, KS 66115
5
The problems of affectively describing and

qualifying low-level data are compounded by the
need for utilizing brief codes in combination with
numerical measures of the analyte in large computer-
ized data bases. These codes, typically supported
by complex definitions, attempt to caution a wide
range of data users with diverse applications. In
attempting to address the data description problem
issues of limit of detection, limit of quantifi-
cation (alternately called limit of quantitation),
criteria of detection, applicable confidence limits,
etc. - must a l l be addressed. Frequently, statisti-
cal rigor is in conflict with the pragmatic use of
the data. Furthermore, the difference between the
quantity of analyte and results of the measuring
system is often confused. The panel is composed of
members of a task force formed to resolve data
coding policies for one of the largest environmental
data systems, STORET. They discuss statistical and
analytical approaches and issues involved in coding
data with the accompanying advantages and hazards.
Overview and I n t r o d u c t i o n ( G e r a r d McKenna)
As more and more e n v i r o n m e n t a l measurements a r e g e n e r a t e d , t h e r e

becomes an o b v i o u s need t o i n p u t e n v i r o n m e n t a l measurements i n t o
computerized databases. T h i s r e p l a c e s t h e a n t i q u a t e d and d i -
a p p e a r i n g p r a c t i c e o f s t u f f i n g t h e d a t a i n t o desk drawers and f i l e
0097-6156/88/0361 -0317$06.00/0
1988 A m e r i c a n C h e m i c a l Society

cabinets. W i t h i n the EPA, t h e r e have a r i s e n a number o f such d a t a

systems such a s , f o r example, STORE, SAROAD, and ODES.
One t r a d e - o f f o f these systems, which c e r t a i n l y c o n t a i n many
p o s i t i v e b e n e f i t s , i s t h a t t h e y d e c r e a s e g r e a t l y the a b i l i t y o f a
c h e m i s t t o communi.cate t o a d a t a user a l l the r e s e r v a t i o n s
he has about h i s measurements. Most o f t h i s a r e a o f c o n c e r n s u r
rounds measurements t h a t l i e v e r y c l o s e t o z e r o . Computer formats
t r y t o r e d u c e t h e c h e m i s t ' s words o f warning t o a l p h a b e t i c a l
symbols, used as codes f o r d i f f e r e n t c l a s s e s o f l i m i t a t i o n s .
A number o f e f f o r t s have t a k e n p l a c e t o e s t a b l i s h c o n v e n t i o n s
of r e p o r t i n g codes. Ihese i n c l u d e , f o r example, work done by t h e
American C h e m i c a l S o c i e t y Committee on E n v i r o n m e n t a l Improvement,
work done by a sub-committee o f t h e American S o c i e t y f o r T e s t i n g
and M a t e r i a l s , t h e I n t e r n a t i o n a l J o i n t Commission between the
u n i t e d S t a t e s and Canada, c o n v e n t i o n s e s t a b l i s h e d by t h e EPA con
t r a c t l a b o r a t o r y progra d convention bein established b th
STORE low l e v e l work group
conventions set at grass-root
p r i v a t e water t e s t i n g l a b o r a t o r i e s .
Out o f t h e s e e f f o r t s have come a m u l t i t u d e o f terms such as
L i m i t o f D e t e c t i o n , Method D e t e c t i o n L i m i t , Instrument D e t e c t i o n
L i m i t , L i m i t o f Q u a n t i t a t i o n , C r i t e r i o n o f D e t e c t i o n and a m u l t i
tude o f symbols such as the l e s s t h a n s i g n , ND, TR ( f o r t r a c e ) , U,
M, J , and W, and K. EPA i s now p r o p o s i n g LTL ( l e s s than lower
l i m i t o f d e t e c t i o n ) and LTC ( l e s s than c r i t e r i a o f d e t e c t i o n ) i n
computer s t a n d a r d s . Some o f the c o n v e n t i o n s come w i t h r i g i d
d e f i n i t i o n s f o r p r e s c r i b e d use and o t h e r s come w i t h vague
d e f i n i t i o n s and a l l o w f o r " a n a l y t i c a l judgment" and f l e x i b i l i t y .
Lack o f c o n v e n t i o n and s t a n d a r d i z a t i o n o f codes and symbols
has caused a g r e a t d e a l o f c o n f u s i o n and i m p a t i e n c e w i t h a n a l y t i
c a l c h e m i s t s from t h e o u t s i d e w o r l d . A q u i c k , s i m p l i s t i c view of
t h e s i t u a t i o n can e r r o n e o u s l y l e a d one t o t h i n k an a g r e e d upon,
s t a n d a r d i z e d c o n v e n t i o n and one t h a t i s u n i v e r s a l l y a c c e p t e d i s n o t
that d i f f i c u l t a task.
To t h e computer p e r s o n t h e r e p r o b a b l y appears the p r o l i f e r a
t i o n o f t o o many c o d e s . To t h e a t t o r n e y , i t appears t h a t a n a l y t i
c a l c h e m i s t s a r e t r y i n g t o " w a f f l e " t o o much. To t h e s t a t i s t i c i a n ,
i t p r o b a b l y appears t h a t t o o much tampering o f d a t a i s g o i n g on and
t o the p u b l i c i t appears t h a t t h e r e may be an attempt t o mask o r
c o n c e a l i n f o r m a t i o n w i t h the use o f t h e s e c o d e s .
Most a n a l y t i c a l c h e m i s t s w i l l q u i c k l y agree on t h e r e a l s o l u
t i o n t o the p r o b l e m . That i s t o choose up f r o n t , o r i f need be,
d e v e l o p a more s e n s i t i v e and a p p r o p r i a t e methodology f o r the
intended use. Such a method s h o u l d get you w e l l up on t h e
measurement and q u a n t i t a t i o n range and l e a v e a l l the low l e v e l
c o d i n g problems b e h i n d . However, t h i s u s u a l l y i s n o t t h e s o l u t i o n .
F i r s t o f a l l , many o f our s t a t e - o f - t h e a r t methods which we have
now a r e n o t more s e n s i t i v e t h a n a r e a r e a s o f e n v i r o n m e n t a l c o n c e r n
and i t t a k e s c o n s i d e r a b l e time and e f f o r t t o d r i v e down the l e v e l s
o f d e t e c t i o n . A l s o , f o r some p a r a m e t e r s , e n v i r o n m e n t a l t o x i c i t y
exposure l e v e l s a r e b e i n g d e t e r m i n e d c o n c u r r e n t l y w i t h s u r v e y
measurements b e i n g made and, t h e r e f o r e , a l l p o s i t i v e f i n d i n g i s
t r e a t e d w i t h g r e a t i n t e r e s t , even i f around z e r o . A l s o , t h e p u b l i c
sometimes p e r c e i v e s any f i n d i n g , even i f q u a l i f i e d , as s i g n i f i c a n t
even as Avogadro's number i s approached.

17. BROSSMAN ET AL. Reporting Low-Level Data 319
C o n s e q u e n t l y , e f f o r t s must c o n t i n u e t o e s t a b l i s h a u n i v e r s a l
convention to t r e a t l o w - l e v e l data. Such c o n v e n t i o n must:
1. be s a t i s f a c t o r y t o t h e a n a l y t i c a l c h e m i s t and n o t compromise o r
o v e r - g e n e r a l i z e t h e l i m i t a t i o n he i s t r y i n g t o communicate;
2. a l l o w f o r v a r y i n g l e v e l s o f r i s k f o r Type I a n d I I e r r o r s t o
accomodate d i f f e r e n t i n t e n d e d uses o f t h e d a t a and d i f f e r e n t
management p h i l o s o p h i e s w i t h r e g a r d t o i t s u s e .
3. be c o n s i s t e n t l y u n d e r s t o o d , even by c h e m i s t s n o t w e l l v e r s e d
i n s t a t i s t i c s so c o n s i s t e n t use i s f o s t e r e d ; and,
4 . be f r e e from m o d i f y i n g o r a d u l t e r a t i n g the d a t a t o make i t
unuseable t o a s t a t i s t i c i a n .
In any e v e n t , coming up w i t h such a c o n v e n t i o n w i l l never p r e c l u d e
t h e need f o r a s e r i o u s i n v e s t i g a t o r o r d a t a u s e r from g o i n g b e h i n d
t h e d a t a i n t h e d a t a base. T h i s i n c l u d e s f u r t h e r communicating
with the o r i g i n a t i n g a n a l y t i c a l chemist e i t h e r d i r e c t l y o r i n -
d i r e c t l y through t h e e x a m i n a t i o
c o n t r o l documentation.
the context of the experimenta
vfe hope t h a t t h e f o l l o w i n g p r e s e n t a t i o n s and your i n p u t will
p r o v i d e us w i t h some h e l p i n r e s o l u t i o n o f t h e s e p r o b l e m s .
A R e t h i n k o f t h e F a c t o r s I n v o l v e d i n R e p o r t i n g R e s u l t s Below t h e
Method D e t e c t i o n L i m i t . (Donald E. King)
The a p p l i c a t i o n o f t h e c o n c e p t o f " d e t e c t i o n l i m i t " as a c r i t e r i o n

by t h e a n a l y t i c a l s c i e n t i s t f o r w i t h h o l d i n g t h e r e s u l t s o f low-
l e v e l measurement i s not s u p p o r t e d . T h i s usage may a r i s e from t h e
mistaken b e l i e f t h a t the o n e - t a i l e d s t a t i s t i c a l t - t e s t i s a t e s t
o f t h e q u a l i t y o f t h e r e s u l t , r a t h e r t h a n t h e e x t e n t t o which t h e
r e s u l t i n d i c a t e s t h a t a n a l y t e i s p r e s e n t i n t h e sample. I t i s also
t h e r e s u l t o f c o n f u s i n g t h e l i m i t a t i o n s o f d e t e c t i o n and measure-
ment w i t h t h e l i m i t a t i o n s o f t h e a n a l y t i c a l p r o c e s s and t h e impact
o f sample m a t r i x e f f e c t s . A f u r t h e r argument a g a i n s t t h e use o f
t h i s c o n c e p t as t h e b a s i s f o r not r e p o r t i n g r e s u l t s i s t h e improper
h y p o t h e s i s t h a t a low r e s u l t i s n e c e s s a r i l y d e r i v e d from a p o p u l a -
t i o n o f r e s u l t s w i t h mean z e r o .
I n s t r u m e n t a l d e t e c t i o n l i m i t (IDL) p r o t o c o l s a r e used t o
d e t e r m i n e when an a n a l o g s i g n a l i s s u f f i c i e n t l y d i f f e r e n t from t h e
b a c k g r o u n d n o i s e t o c o n c l u d e t h a t a m e a s u r e a b l e " r e a l " s i g n a l has
been o b s e r v e d . But IDL t e r m i n o l o g y and e s t i m a t i o n p r o t o c o l s a r e
not standard.
More r e c e n t l y d a t a r e p o r t a b i l i t y has been l i n k e d t o t h e
r e p e a t a b i l i t y o f measurement o f samples a f t e r a more o r l e s s ex-
t e n s i v e sample p r e p a r a t i o n and a n a l y s i s p r o c e s s . Many a n a l y t i c a l
c h e m i s t s c u r r e n t l y c o n s i d e r i t improper t o r e p o r t measurements
o b s e r v e d t o be below t h e i r Method D e t e c t i o n L i m i t (MDL). They
f a i l t o r e c o g n i z e t h a t t h e measurement was i n f a c t above t h e IDL
and i s t h e r e f o r e a r e a l r e s u l t .
The a n a l y s t o f t e n tends t o o v e r - e s t i m a t e t h e " d e t e c t i o n
l i m i t " b e c a u s e o f c o n c e r n about t h e a c c u r a c y o f s e t t i n g z e r o , t h e
p o t e n t i a l f o r b i a s , m i s i d e n t i f i c a t i o n , e t c . , and f e a r t h a t t h e d a t a
u s e r w i l l draw u n s u p p o r t a b l e c o n c l u s i o n s .

The symbol "<" ( l e s s than) has been v a r i o u s l y used t o i n d i

c a t e a z e r o r e s u l t , o r t o i n d i c a t e a measurement below t h e " d e t e c
t i o n l i m i t " but not n e c e s s a r i l y " z e r o " . More r e c e n t l y , STORET has
p r o v i d e d the codes W, U, T, and M to i n d i c a t e when a measurable r e
sponse v/as o b s e r v e d and whether o r not a r e s u l t has been r e p o r t e d .
"W" s i g n i f i e s a n o n - r e s p o n s e . "T" s i g n i f i e s a measured r e s p o n s e
below 1.64 SO w h i c h has been r e p o r t e d . "U" s i g n i f i e s a r e s u l t be
low MDL which has not been r e p o r t e d . "M" s i g n i f i e s a r e s u l t be
tween MDL and LOQ.
I t i s presumed t h a t MDL i s s e t a t 3SD and LOQ a t 10 SD, b u t
the t y p e o f d a t a and p r o t o c o l used to e s t i m a t e SD i s n o t v e r i f i
able. A n a l y s t s have a tendency t o i n f l a t e t h e i r e s t i m a t e s o f MDL.
The use of W and f o r c o d i n g l o w - l e v e l r e s u l t s was i n c l u d e d i n
s t a n d a r d D4210-83 p r e p a r e d by t h e A.S.T.M. D19 Committee (1)
In t h e L a b o r a t o r y S e r v i c e s Branch o f t h e O n t a r i o M i n i s t r y o f
t h e Environment, t h e code <W d <T introduced l
ago. Data c o d e d <W i s
Data coded <T i s t o be
zero measurements. R e c e n t l y t h e v a l u e o f W i s s e t e q u a l t o t h e
" L o g i c i a l R e a d i n g Increment" ( L R I ) . The v a l u e o f i s u s u a l l y
5 times W. L a r g e r f a c t o r s o f 10 o r 2 0 t i m e s W a r e used f o r methods
subject to greater uncertainty i n target analyte i d e n t i f i c a t i o n or
q u a n t i t a t i o n . Because W i s u s u a l l y a p p r o x i m a t e l y 2/3 o f the
s t a n d a r d d e v i a t i o n , (=5W) w i l l exceed t h e 95% c o n f i d e n c e l e v e l
f o r d e t e c t i o n o f a n a l y t e i n the sample, and w i l l o f t e n be e s s e n
t i a l l y e q u i v a l e n t t o t h e 3s d e t e c t i o n l i m i t .
S t a n d a r d d e v i a t i o n (SD) i s t h e o n l y l o g i c a l b a s i s f o r
s e l e c t i n g a r e a d i n g i n c r e m e n t ( R I ) . In o r d e r t o o b t a i n a r e l i a b l e
e s t i m a t e o f SD, t h e d a t a used f o r i t s c a l c u l a t i o n s h o u l d have been
r e p o r t e d i n i n c r e m e n t s a t l e a s t s m a l l e r than 2/3 t h e SD. While
t h e r e i s no p a r t i c u l a r b e n e f i t i n r e a d i n g i n i n c r e m e n t s s m a l l e r
t h a n o n e - h a l f o f SD, t h e v e r y h i g h r i s k o f e r r o r a s s o c i a t e d w i t h
e x c e s s i v e r o u n d o f f i s r e a d i l y d e m o n s t r a t e d {2).
Thus, t h e b a s i s f o r s e l e c t i n g r e a d i n g i n c r e m e n t s s h o u l d be t o
p i c k a d i g i t 1, 2, o r 5 ( w i t h a p p r o p r i a t e d e c i m a l l o c a t i o n ) , which
i s j u s t l e s s t h a n the e s t i m a t e d SD. I f SD i s i n the range 0.5 t o
1.0, t h e L o g i c a l Reading Increment (LRI) s h o u l d be no l a r g e r t h a n
0.5. But s m a l l e r r e a d i n g and r e p o r t i n g i n c r e m e n t s a r e p e r m i t t e d .
The p r o c e s s o f s t a t i s t i c a l i n f e r e n c e r e q u i r e s us t o s e l e c t an
h y p o t h e s i s ( f a n c y word f o r assumption) about the r e s u l t , and then
prove t h a t t h i s h y p o t h e s i s was i n c o r r e c t . In t h e c a s e o f Method
D e t e c t i o n L i m i t , we assume t h a t the r e s u l t b e l o n g s t o a d i s t r i b u
t i o n whose mean i s c e n t e r e d on " z e r o " . Based on a o n e - t a i l e d
t e s t , i f t h e a n a l y t i c a l r e s u l t i s f a r enough away from z e r o we
then c o n c l u d e t h a t such a r e s u l t must come from some o t h e r d i s t r i
b u t i o n , i n which c a s e t h e r e i s some l i k e l i h o o d t h a t the sample
c o n t a i n s the t a r g e t a n a l y t e . MDL i s c o n v e n t i o n a l l y s e t a t 3 SD.
I f a sample a c t u a l l y c o n t a i n e d an amount e q u a l t o t h e MDL,
t h e n 50% o f the time a r e s u l t would be o b t a i n e d which was below
MDL. F a i l u r e t o r e p o r t t h a t r e s u l t can i n d u c e a " f a l s e n e g a t i v e "
decision.

T h e r e i s no l o g i c a l b a s i s f o r t h e i n i t i a l h y p o t h e s i s t h a t a
low r e s u l t comes from a p o p u l a t i o n w i t h mean z e r o . A low r e s u l t
can be o b t a i n e d from any o f t h e b i l l i o n s o f the r e s u l t p o p u l a t i o n s
on the a n a l o g number l i n e between z e r o and t w i c e t h e d e t e c t i o n
limit. B e f o r e the s t a t i s t i c a l t e s t i s even a p p l i e d , t h e r e i s
a l r e a d y an immeasurable r i s k o f a " f a l s e n e g a t i v e " c o n c l u s i o n ,
a g g r a v a t e d by t h e f a c t t h a t c o n v e n t i o n a l wisdom r e j e c t s the r e
p o r t i n g o f r e s u l t s below MDL.
A more a p p r o p r i a t e use o f t h e o n e - t a i l e d t e s t f o l l o w s . I f
a h e a l t h g u i d e l i n e were s e t a t 10 u n i t s , and t h e SD were about
3.3 ( i . e . , MDL was 10), t h e n p r o b a b i l i t y t a b l e s p r e d i c t t h a t a
r e s u l t g r e a t e r t h a n 14 ( i . e . , 10 + 1.64 3.3) would s u g g e s t t h e
sample c o n t a i n s more than t h e g u i d e l i n e ( r i s k o f e r r o r <5%),
w h i l e a r e s u l t o f l e s s t h a n 6 would s u g g e s t the sample c o n t a i n s
l e s s t h a n the g u i d e l i n e . N o t i c e t h a t f a i l u r e to r e p o r t these
low v a l u e s p r e v e n t s d e c i s i o making
The a n a l y s t does
g i v e n a v e r y low r e s u l t , when p r e v i o u s e x p e r i e n c e i n d i c a t e s t h a t
v e r y few, i f any, s i m i l a r samples from a s i m i l a r s o u r c e have
y i e l d e d any s o r t o f a r e s p o n s e , l e t a l o n e a t r a c e . But i f many
s i m i l a r samples y i e l d low b u t non-zero r e s u l t s , he must admit a t
some p o i n t t h a t a low r e s u l t may mean " p r e s e n c e " . We have found
f o r some p a r a m e t e r s t h a t a new more s e n s i t i v e method s t i l l f a i l s
to d e t e c t a n a l y t e , w h i l e f o r o t h e r s t h e p r e v i o u s i n d i c a t i o n s o f
presence are confirmed. T h i s i s j u s t the s o r t of evidence t h a t
s t a t i s t i c i a n s l o o k f o r , and which i s d e n i e d t o t h e d a t a u s e r when
we f a i l t o r e p o r t low r e s u l t s .
The a n a l y s t i s o b l i g e d t o r e p o r t what was o b s e r v e d as w e l l

as h i s c o n c l u s i o n . I f the s i g n a l was caused by the t a r g e t e d
a n a l y t e , i t must be q u a n t i t a t e d . An i m p o r t a n t f a c t o r i n r e v i e w i n g
our p o s i t i o n on the r e p o r t i n g o f low d a t a i s t o r e c o g n i z e t h e
e f f e c t o f the c o m p u t e r i z e d d a t a b a s e s i n s e p a r a t i n g t h e d a t a
s o u r c e from t h e d a t a u s e r . I f we do not r e p o r t t h e r e s u l t , we a r e
t h e n r e p o r t i n g an i n f e r e n c e based on a s i n g l e v a l u e .
I t i s c e r t a i n l y i m p o r t a n t t o warn d a t a u s e r ' s t h a t some o f
t h e i r d a t a was o b t a i n e d from t h e bottom o f t h e a n a l y t i c a l range.
R e g a r d l e s s o f t h e p r e c i s i o n o f a n a l y s i s and measurement, t h e
a b i l i t y o f t h e a n a l y s t t o d e f i n e a n a l y t e z e r o and t o d e t e r m i n e and
c o r r e c t l y a p p l y a l l t h e c o r r e c t i o n s f o r b a c k g r o u n d , r e a g e n t , and
o t h e r l a b b l a n k s , i s s t i l l l i m i t e d by t h e a n a l y t e , t h e sample
m a t r i x , t h e a n a l y t i c a l method, and h i s o r h e r s k i l l and
experience.
T h i s p o t e n t i a l f o r e r r o r does n o t mean t h a t low r e s u l t s a r e
necessarily i n error. The mechanics o f e s t i m a t i n g and s e t t i n g
z e r o and d i g i t i z i n g r e a d i n g s have improved d r a m a t i c a l l y i n t h e
l a s t few y e a r s . P o t e n t i a l l y , i n a b s o l u t e terms, t h i s data i s
much more a c c u r a t e than p r e v i o u s l y c o n s i d e r e d as l o n g as z e r o
calibration i s controlled. Q u a l i f y i n g low r e s u l t s a l l o w s t h e

a n a l y s t to c r o s s h i s / h e r f i n g e r s , while a l l o w i n g the data user

f u l l access to a l l a v a i l a b l e information.
Reference:
1. ASTM - D19 D4210-33, I n t r a l a b o r a t o r y Q u a l i t y C o n t r o l and a

D i s c u s s i o n on R e p o r t i n g Low-Level Data, A
S t a n d a r d P r a c t i c e (1983).
2. In p r e p . D.E. K i n g , A R e t h i n k o f t h e F a c t o r s I n v o l v e d i n
M e a s u r i n g , R e p o r t i n g , and I n t e r p r e t i n g Data.
R e p o r t i n g o f Low-Level Data f o r a S p e c i f i c C o m p u t e r i z e d Database

(Robert K l e o p f e r )
S i n c e 1982, t h e U.S. E n v i r o n m e n t a l P r o t e c t i o n Agency L a b o r a t o r y i n

Kansas C i t y has been a c t i v e l y i n v o l v e d i n i n v e s t i g a t i n g s i t e s con-
taminated with " d i o x i n "
have been a n a l y z e d f o r
A computer database was d e v e l o p e d t o t r a c k a number o f d i f f e r e n t
t y p e s o f i n f o r m a t i o n p e r t a i n i n g t o each sample, i n c l u d i n g sample
i d e n t i f i c a t i o n number, a n a l y t i c a l r e s u l t s and q u a l i f i c a t i o n code.
A "U" code was used t o s i g n i f y t h a t TCDD was not d e t e c t e d i n
a g i v e n sample. The d e t e c t i o n l i m i t s were c a l c u l a t e d f o r each
i n d i v i d u a l sample based on a 2-1/2 s i g n a l t o n o i s e d e f i n i t i o n f o r
t h e p r i m a r y i o n s r e c o r d e d by t h e Gas Chromatograph/Mass S p e c t r o -
m e t e r / D a t a System (GC/MS/DS) An i s o t o p e d i l u t i o n method based on
i s o t o p i c a l l y l a b e l e d d i o x i n was used t o c o n v e r t s i g n a l l e v e l t o
concentration units. Thus t h e a c t u a l r e p o r t e d d e t e c t i o n l i m i t
v a r i e d f r o m sample t o sample due t o d i f f e r e n c e s i n i n t e r n a l s t a n -
d a r d r e c o v e r y , c h e m i c a l n o i s e , and i n s t r u m e n t a l s e n s i t i v i t y . The
"U" code was used when the s i g n a l t o n o i s e was l e s s t h a n 2-1/2 or
the q u a l i t a t i v e i d e n t i f i c a t i o n c r i t e r i a f o r d i o x i n were n o t
achieved. F o r example, t h e "U" code was used i n i n s t a n c e s where
s i g n a l was p r e s e n t i n e x c e s s o f 2-1/2 t i m e s t h e n o i s e , b u t t h e
i s o t o p e r a t i o s (due t o n a t u r a l l y o c c u r r i n g c h l o r i n e 35 and
c h l o r i n e 3 7 i n TCDD) were o u t s i d e o f a c c e p t a b l e r a n g e s .
Data q u a l i t y was i n d i c a t e d w i t h t h r e e d i f f e r e n t c o d e s . A "V"
code was used t o i n d i c a t e c o m p l e t e l y v a l i d d a t a . Those d a t a met
s t r i n g e n t EPA c r i t e r i a i n c l u d i n g t h e c o r r e c t measurement o f p e r -
formance samples. An " I " code was used t o i n d i c a t e i n v a l i d d a t a .
Those d a t a were c o n s i d e r e d c o m p l e t e l y unusable and i n d i c a t e d t h a t
a " c r i t i c a l " c o n t r o l f e a t u r e was n o t a c h i e v e d . F o r example, t h e
r e p o r t i n g o f a "not d e t e c t e d " f o r a p o s i t i v e p e r f o r m a n c e sample
would i n v a l i d a t e a l l a s s o c i a t e d samples. F i n a l l y , a " J " code was
used t o s i g n i f y d a t a i n w h i c h " n o n - c r i t i c a l " c o n t r o l s were n o t
met. F o r example, i f a p e r f o r m a n c e sample was r e p o r t e d t o c o n t a i n
TCDD a t l e v e l s j u s t o u t s i d e o f the a c c e p t a n c e window, then a s s o -
c i a t e d samples were coded " J " . The d a t a were c o n s i d e r e d u s e a b l e
from a q u a l i t a t i v e a s p e c t ( d i o x i n was p r e s e n t ) but t h e q u a n t i t a -
t i v e amount was i n q u e s t i o n .

17. BROSSMANETAL. Reporting Low-Level Data
By u s i n g t h e s e s i m p l e d e f i n i t i o n s we were a b l e t o m a i n t a i n a
l a r g e database which c o n t a i n e d the n e c e s s a r y i n f o r m a t i o n w i t h o u t
confusing the data user.
ACS P r i n c i p l e s f o r E n v i r o n m e n t a l A n a l y s i s Recommendations f o r
R e p o r t i n g Low-Level D a t a . (John T a y l o r )
Measurements a r e made f o r many p u r p o s e s b u t o r d i n a r i l y w i t h some

s p e c i f i c use i n mind. Once the use i s d e f i n e d , d a t a q u a l i t y
o b j e c t i v e s can be e s t a b l i s h e d and t h e most c o s t - e f f e c t i v e way t o
a c h i e v e them can be sought. Sometimes t h e d a t a q u a l i t y o b j e c t i v e s
a r e n o t r e a c h e d and t h i s can cause u t i l i z a t i o n p r o b l e m s . When
d a t a a r e u s e d f o r o t h e r than t h e i r i n t e n d e d p u r p o s e , i t s q u a l i t y
a l s o can make such usage i n a d v i s a b l e .
Much o f measurement d a t a can f a l l i n t o two d i s t i n c t c l a s s e s .
In t h e f i r s t c a s e , t h e
o f a p o p u l a t i o n , f o r exampl
fish. Q u a n t i t a t i v e value
s t a n d a r d d e v i a t i o n s and o n l y t r u e o u t l i e r s a r e e x c l u d e d from t h e
d a t a s e t , o f t e n as a f i n a l s t e p i n the d a t a a n a l y s i s . In t h e
second c a s e , the q u e s t i o n may be whether a s p e c i f i e d v a l u e has
been exceeded o r even i f a d e t e c t a b l e amount o f a n a l y t e i s p r e s e n t
Only t h e answer yes o r no i s c o n s i d e r e d t o be i m p o r t a n t and numer-
i c a l v a l u e s may o r may n o t be r e p o r t e d , e x c e p t f o r c r i t i c a l
samples.
In r e c o g n i t i o n o f t h e v a r i o u s u s e s and chances f o r m i s i n t e r -
p r e t a t i o n o f d a t a , t h e ACS Committee on E n v i r o n m e n t a l Improvement
(1) made s e v e r a l recommendations f o r r e p o r t i n g d a t a t h a t may even
seem t o be c o n t r a d i c t o r y when t a k e n o u t o f c o n t e x t .
The r e p o r t i s s u e d by t h e committee recommends t h a t a l l d a t a
r e p o r t e d s h o u l d have c l e a r l y d e f i n e d l i m i t s o f u n c e r t a i n t y . When
t h i s i s done, any u s e r may e v a l u a t e i t s u s e f u l n e s s f o r any p u r p o s e
No discouragement t o r e p o r t n u m e r i c a l v a l u e s f o r any d a t a was made
n o r i n t e n d e d as l o n g as i t s u n c e r t a i n t y i s s p e c i f i e d .
The q u e s t i o n o f d e t e c t i o n , t h a t i s the d e c i s i o n based o n
d a t a , was a d d r e s s e d as f o l l o w s . I f a measured v a l u e i s s m a l l e r
t h a n i t s u n c e r t a i n t y (3 s t a n d a r d d e v i a t i o n s ) t h e measured v a l u e i s
n o t c o n s i d e r e d t o be s i g n i f i c a n t a t t h a t l e v e l o f c o n f i d e n c e ,
hence t h e a n a l y t e sought i s r e p o r t e d a s "not d e t e c t e d " (ND) w i t h
t h e l i m i t o f d e t e c t i o n (3 sigma) g i v e n i n p a r e n t h e s e s . T h i s means
t h a t by chance a l o n e , t h e v a l u e o b t a i n e d c o u l d have been found f o r
a sample n o t c o n t a i n i n g any m e a s u r a b l e amount o f t h e a n a l y t e o f
concern.
The word measurable i s a keyword. O r d i n a r i l y some v a l u e i s
o b t a i n e d when any sample i s a n a l y z e d . But i f t h e r e s u l t i s n o t
s t a t i s t i c a l l y s i g n i f i c a n t , one can have l i t t l e c o n f i d e n c e i n i t .
The "3 sigma" l i m i t i s a r b i t r a r i l y recommended as t h e d e c i s i o n
level.
Because o f t h e l a r g e u n c e r t a i n t y o f d a t a e x c e e d i n g but c l o s e
to t h e above l i m i t , t h e r e p o r t recommends t h a t such i n d i v i d u a l
v a l u e s be r e p o r t e d as d e t e c t e d (D) w i t h l i m i t s o f d e t e c t i o n i n d i -
c a t e d r a t h e r t h a n r e p o r t e d as numbers. The r e a s o n f o r t h i s i s t o
emphasize t h e s e m i - q u a n t i t a t i v e n a t u r e o f such d a t a and t o d i s -
c o u r a g e o v e r - i n t e r p r e t a t i o n by u n s o p h i s t i c a t e d u s e r s . For ex-

ample, t h e r e s u l t 5 micrograms 3 micrograms h a r d l y c a n be

c o n s i d e r e d as r e l i a b l e to one s i g n i f i c a n t f i g u r e i n t h a t t h e
p r o b a b l e range o f 2 micrograms t o 8 micrograms l i m i t s i t s
quantitative usefulness severely.
As a measured v a l u e i n c r e a s e s , i t s r e l a t i v e u n c e r t a i n t y de
c r e a s e s so t h a t a t some p o i n t i t becomes q u a n t i t a t i v e l y u s e f u l .
The lower l i m i t o f q u a n t i t a t i o n (LOQ) was a r b i t r a r i l y d e f i n e d as
"10 sigma". A c c o r d i n g l y , the r e p o r t recommends r e p o r t i n g numer
i c a l v a l u e s g r e a t e r than the LOQ as numbers t o g e t h e r w i t h t h e i r
associated uncertainties.
The above c o n s i d e r a t i o n s a r e b a s e d on i n t e r p r e t i n g a s i n g l e
measurement. I f the mean o f a s e t o f measurements i s r e p o r t e d
o r used f o r d e c i s i o n , t h e n the l i m i t s a r e m u l t i p l i e d by t h e
factor 1/^n.
The LOQ i s u s e f u l f o r d e s c r i b i n g t h e c a p a b i l i t y o f a metho
dology s i n c e i t represent
q u a n t i t a t i v e s i n g l e measurements
t a i n t y o f measurement ( o f t e
r e g u l a t o r s ) t h e r e p o r t s t r e s s e s t h a t r e g u l a t i o n s s h o u l d not be s e t
below a measurable l e v e l , namely t h e LOQ of t h e methodology recom
mended f o r m o n i t o r i n g c o m p l i a n c e . Below such a l e v e l , r e g u l a t i o n s
h a r d l y can be e n f o r c e a b l e due to measurement u n c e r t a i n t y o f random
error.
In t h i n k i n g about the above, one must u n d e r s t a n d the impor
t a n t d i f f e r e n c e s between the two k i n d s o f d a t a d e s c r i b e d . Unfor
t u n a t e l y time and c o s t c o n s i d e r a t i o n s i n f l u e n c e b o t h the ways d a t a
a r e measured and t h e i r q u a l i t y so t h a t d a t a made f o r one p u r p o s e
may n o t have p e e r s t a t u s w i t h s i m i l a r d a t a f o r a n o t h e r p u r p o s e .
T h i s r e l a t e s t o h i s t o r i c a l d a t a , as w e l l . As methodology improves
and as a t t e n t i o n t o q u a l i t y a s s u r a n c e i n c r e a s e s , d a t a once
a c c e p t a b l e may become o b s o l e t e , and may need t o be d i s c a r d e d , r e
g a r d l e s s o f how p a i n f u l t h i s may be. This i s e s p e c i a l l y true of
low-level data. Data r e p o r t e d as " l e s s - t h a n some v a l u e " may have
l i t t l e q u a n t i a t i v e v a l u e a t some l a t e r d a t e even i f a number had
been r e p o r t e d because o f measurement u n c e r t a i n t y . When d a t a a r e
2
p r o p e r l y w e i g h t e d by t h e f a c t o r 1/6 , as i s n e c e s s a r y when combin
ing i t with other data, the r a t i o n a l e f o r d i s c a r d i n g i s c l a r i f i e d .
One can always make q u a n t i t a t i v e use o f any r e l i a b l e d a t a ,
even i f below the d e t e c t i o n l e v e l , but o n l y when i t s u n c e r t a i n t y
i s known and a s u f f i c i e n t number i s a v a i l a b l e . As d e t e c t i o n l e v e l s
are l o w e r e d , l e v e l s o f c o n c e r n a r e l o w e r e d as w e l l , so t h a t
a n a l y s t s a r e c o n t i n u a l l y a s k e d t o "measure the immeasurable".
What can be done i n r e s e a r c h s i t u a t i o n s o f t e n i s i n f e a s i b l e f o r
routine decisions. A c c o r d i n g l y , t h e r e w i l l always be a g u l f be
tween " r e s e a r c h " and " p r a c t i c a l " d a t a . The ACS recommendations
were d i r e c t e d l a r g e l y a t the l a t t e r type. W h i l e t h e recommenda
t i o n s may appear t o encourage l o s s o f q u a n t i t a t i v e i n f o r m a t i o n ,
the a c t u a l l o s s may have l i t t l e s i g n i f i c a n c e i n most c a s e s due t o
quality considerations. However, no d a t a s h o u l d be e d i t e d i n t h i s
o r any o t h e r manner i f s o - d o i n g would p r e c l u d e i t s f u t u r e use f o r
an o t h e r - w i s e u s e f u l p u r p o s e .
When e v a l u a t i n g d a t a s e t s , o b j e c t i v e l y , one i s f o r c e d t o con
c l u d e t h a t much o f i t i s b e i n g p r o d u c e d by l a b o r a t o r i e s w i t h
l i m i t e d competence t o do so. In f a c t , i t would be b e t t e r not t o
r e p o r t some d a t a . C o n t r a c t s t o p r o v i d e d a t a o f t e n a r e l e t on t h e

b a s i s o f p r i c e which has no r e l a t i o n t o q u a l i t y . Too many d a t a

a r e a c c e p t e d on t h e b a s i s o f f a c e v a l u e . There i s a need t o
f i l t e r d a t a f o r i t s q u a l i t y b e f o r e use and to r e j e c t i t i f need
be. A means t o q u a l i f y l a b o r a t o r i e s and t o m o n i t o r c o n t i n u i n g
p e r f o r m a n c e i s n e c e s s a r y i f t h i s problem i s t o be s o l v e d .
Data, j u s t as any commodity, must be bought on t h e b a s i s o f
quality. I f i t does not meet p r e - d e s c r i b e d q u a l i t y s t a n d a r d s , i t
i s not p a i d f o r and r e j e c t e d , j u s t as i n t h e c a s e o f goods and
p r o d u c t s (2).
References
1. Principles of Environmental Analysis, Anal. Chem. Vol. 55,

2210-18 (1983).
2. J. K. Taylor, Quality Assurance for a Measurement Program, in
"Environmental Samplin fo Hazardou Wastes" G E Schweitze
and J. A. Santolucito
Washington, D.C. 2003
Summary and C l o s u r e ( M a r t i n W. Brossman)
Our p a n e l i s t s have d i s c u s s e d a range o f approaches and i s s u e s

r e l a t e d t o the c o d i n g and r e p o r t i n g o f l o w - l e v e l d a t a f o r
computerized data bases. The v e r y d i v e r s i t y o f t h e d i s c u s s i o n
i l l u s t r a t e s t h e d i f f i c u l t y we and o t h e r s have had i n a r r i v i n g
a t a consensus on a m u t u a l l y a g r e e d upon a p p r o a c h . We a r e f a c e d
with a very s e r i o u s , r a t h e r i n t r a c t a b l e , s i t u a t i o n . Many
p o l l u t a n t s a r e h a z a r d o u s down t o l e v e l s which t a x our a b i l i t y t o
measure t h e i r p r e s e n c e and q u a n t i t y w i t h any r e l i a b i l i t y . However,
t h e r e a r e no commonly a g r e e d upon methods t o d e s c r i b e the r e s u l t s
o f t h e s e l o w - l e v e l measurements and even l e s s c o n s e n s u s on ways t o
reduce t h e d e s c r i p t i o n s t o codes n e c e s s a r y t o use w i t h our l a r g e
computerized data bases. (We even f i n d s t r o n g d i f f e r e n c e s i n
p r e f e r e n c e o r a c c e p t a n c e o f the term q u a n t i t a t i o n f o r q u a n t i f i -
cation. ) A b r i e f summary o f our p a n e l d i s c u s s i o n s w i l l i l l u s t r a t e
some o f t h e problems and p o s s i b l e s o l u t i o n s . G e r a l d McKenna has
d i s c u s s e d t h e b a s i c i s s u e s i n v o l v e d w i t h l o w - l e v e l measurement and
coding. He has l i s t e d f o u r c r i t e r i a which a u n i v e r s a l c o n v e n t i o n
f o r t r e a t i n g l o w - l e v e l d a t a s h o u l d meet. W h i l e n o t a l l i n c l u s i v e ,
t h e s e v e r y c r i t e r i a i l l u s t r a t e t h e immensity o f the t a s k . The
c r i t e r i a r e q u i r e s a t i s f y i n g b o t h the a n a l y t i c a l c h e m i s t and a range
of i n t e n d e d d a t a u s e r s - w h i l e s i m u l t a n e o u s l y m e e t i n g t h e r i g o r o f
t h e s t a t i s t i c i a n and be u n d e r s t a n d a b l e and u s e a b l e f o r t h e non-
statistician. Such r e q u i r e m e n t s s u g g e s t t h a t a t b e s t we may have
t o be c o n t e n t w i t h a s e r i e s o f c o n v e n t i o n s o f r e s t r i c t e d a p p l i -
c a t i o n f o r some t i m e .
The K i n g , K l e o p f e r and T a y l o r p a p e r s i l l u s t r a t e a v a r i e t y
o f c o n v e n t i o n s and i s s u e s r e l a t e d t o d e s c r i b i n g l o w - l e v e l d a t a .
(Henry Kahn's d i s c u s s i o n o f s t a t i s t i c a l i s s u e s was u n a v a i l a b l e
for i n c l u s i o n here.) Don K i n g d i s c u s s e s problems r e l a t e d t o
terms and c o d i n g c o n v e n t i o n s t o g e t h e r w i t h some b a s i c s t a t i s t i -
cal issues. Don K i n g ' s p r e f e r e n c e i s f o r a v a r i a t i o n o f t h e "W"
and "T" codes p r o p o s e d by ASTM. He argues f o r t h e importance
of c a r e f u l e s t a b l i s h m e n t o f such l i m i t s as method d e t e c t i o n
and i n s t r u m e n t d e t e c t i o n l i m i t s w h i l e a t t h e same time

a c c u r a t e l y r e p o r t i n g f i n d i n g s beyond the l i m i t s , w i t h q u a l i f i -
cation. " Q u a l i f y i n g low r e s u l t s a l l o w s the a n a l y s t t o c r o s s
h i s / h e r f i n g e r s , while a l l o w i n g the data user f u l l access to a l l
available information."
Robert K l e o p f e r d i s c u s s e s a d i f f e r e n t a s p e c t o f r e p o r t i n g o f
l o w - l e v e l d a t a f o r c o m p u t e r i z e d d a t a b a s e s t h a n t h a t d i s c u s s e d by
Don K i n g o r J o h n T a y l o r . K l e o p f e r discusses a computerized data
system and c o d i n g a p p r o a c h d e s i g n e d f o r a s i n g l e a n a l y t e - D i o x i n .
In t h e c a s e o f D i o x i n , a comprehensive s e t o f c o n t r o l s and
p e r f o r m a n c e r e q u i r e m e n t s have been e s t a b l i s h e d . These c o n t r o l s and
performance requirements are i n c l u d e d i n a data review p r o t o c o l .
Data may be r e j e c t e d o r a c c e p t e d b a s e d on c o n d i t i o n s r e l a t e d t o a
s i n g l e measurement o r a d a t a s e t . The d a t a r e v i e w e r , u s i n g the
d a t a r e v i e w p r o t o c o l , w i l l a s s i g n a "V" code t o i n d i c a t e the d a t a
meets c o n t r o l and p e r f o r m a n c e r e q u i r e m e n t s as s p e c i f i e d i n the d a t a
review p r o t o c o l . An " I " cod indicate i n v a l i d dat d "J"
code i n d i c a t e s some n o n c r i t i c a
codes and c o n v e n t i o n s d i s c u s s e
a p p r o a c h g e t s i n t o the i s s u e s o f d a t a v a l i d i t y b o t h a t t h e " l i m i t s
o f d e t e c t i o n " and t h o s e r e l a t e d t o c r i t e r i a a p p l i c a b l e a t o r d i n a r y
l e v e l s of a n a l y s i s .
J o h n T a y l o r f o c u s e s p r i m a r i l y on i s s u e s r e l a t e d t o the
recommendations o f t h e ACS Committee on E n v i r o n m e n t a l Improvement,
which he c o - a u t h o r e d . The c o n v e n t i o n s recommended h e r e a r e
d e s i g n e d t o e n s u r e d a t a c e r t a i n t y f o r use i n r o u t i n e d e c i s i o n s a s
opposed t o r e s e a r c h use. L i m i t s o f d e t e c t i o n a r e s e t a t 3 sigma
l i m i t s and l i m i t s o f q u a n t i t a t i o n a t 10 sigma l i m i t s . Taylor notes
t h a t a g u l f w i l l always e x i s t between r e s e a r c h and " p r a c t i c a l "
d a t a - and we may a l s o add t h e i m p l i e d need w i l l t h e r e f o r e e x i s t
f o r d i f f e r e n t c o n v e n t i o n s and r e s u l t i n g n o n - s t a n d a r d i z a t i o n .
The ACS a p p r o a c h d i s c u s s e d by John T a y l o r and e x t e n s i o n s o f
t h e ASTM a p p r o a c h e s d i s c u s s e d by Don K i n g i l l u s t r a t e some o f t h e
p r a c t i c a l d i f f i c u l t i e s f a c e d i n a t t e m p t i n g t o compare and evaluate
a p p r o a c h e s . The ASTM a p p r o a c h u t i l i z e s terms, c r i t e r i a o f d e t e c -
t i o n and l i m i t o f d e t e c t i o n . The ASC recommendations i n c l u d e
t e r m s , l i m i t o f d e t e c t i o n and l i m i t o f q u a n t i t a t i o n . Both
approaches p r o v i d e a s t a t i s t i c a l b a s i s f o r t h e t e r m s . However,
a t t e m p t s by our t a s k f o r c e on l o w - l e v e l d a t a t o make a r i g o r o u s
c o n c e p t u a l and s t a t i s t i c a l c o m p a r i s i o n o f the approaches have been
unsuccessful. Even s i m i l a r terms a r e d e f i n e d i n d i f f e r e n t , non-
comparable ways, and a d d i t i o n a l terms and c o n c e p t s a r e u s e d which
a r e unique t o each a p p r o a c h .
We have a t t e m p t e d t o d e s c r i b e i s s u e s r e l a t e d t o and
approaches f o r comprehensive and r e l i a b l e r e p o r t i n g o f l o w - l e v e l
data f o r computerized data bases. H o p e f u l l y , we have a l s o conveyed
some c o n c e p t o f t h e importance o f the p r o b l e m and t h e d i v e r s i t y o f
a p p r o a c h e s aimed a t problem r e s o l u t i o n . However, we f i n d no a g r e e d
upon methods t o d e s c r i b e the r e s u l t s o f l o w - l e v e l a n a l y t e measure-
ments o r consensus on r e d u c i n g d e s c r i p t i o n s t o codes f o r use i n
l a r g e computerized data bases. P a r t o f t h e d i f f i c u l t y may be
a s c r i b e d t o c o n f l i c t i n g demands i n c l u d i n g m e e t i n g t h e needs o f
diverse users. However, t h e r e a l s o a p p e a r s t o be a l a c k of r i g o r

i n a d d r e s s i n g t h e problem even where a l o w - l e v e l d a t a c o n v e n t i o n

and i t s codes a r e d e s i g n e d f o r a h i g h l y s p e c i f i c d a t a b a s e . A
s y s t e m a t i c e x a m i n a t i o n and s p e c i f i c a t i o n o f t h e d a t a need and use
f o l l o w e d by a c a r e f u l development o f t h e a p p r o p r i a t e " c o n v e n t i o n "
and codes c o u l d be h e l p f u l .
C e r t a i n l y , a c a r e f u l l y d e v e l o p e d and documented p r o t o c o l
would p e r m i t more e f f e c t i v e e v a l u a t i o n o f t h e v a l i d i t y and
appropriateness of conventions f o r d e s c r i b i n g l o w - l e v e l data.
A j o i n t e f f o r t o f ACS and ASTM c o u l d , a t a minimum, recommend t h e
i s s u e s t o be a d d r e s s e d and a d o c u m e n t a t i o n s t r u c t u r e f o r l o w - l e v e l
data conventions. Our t a s k f o r c e f o r one would be p l e a s e d t o
p r o v i d e i n p u t s r a n g i n g from g e n e r a l f r u s t r a t i o n s t o s p e c i f i c
recommendations.
R E C E I V E D M a r c h 2, 1987

Author Index
Bauer, C F . , 194 Liggett, Walters., 210

Brossman, Martin W., 317 McCormack, M i k e , 64
Crummett, Warren B., 288 M c K e n n a , Gerard, 317
Currie, L l o y d ., 1,171 Mellor, R . ., 244
Epstein, Michael S., 109 Midkiff, Charles R., Jr., 288
Grant, C . L . , 194 Moss, Thomas H . , 70
Hallowell, S. F., 224 Owens, K . G . , 194
Hanneman, W. William, 288 Parr, Robert M , 171
Hare, P. E . , 275 Reutter, D . J., 224
Harrington, C L . 244 Rogers L B. 94
Horwitz, William
Johnson, Jame
Johnson, Janet ., 266 St. John, P. ., 275
K a h n , Henry, 317 Steigstra, H . , 126
K i n g , Donald, 317 Sturdivan, L a r r y , 288
K i r c h m e r , C l i f f J., 78 Taylor, John K . , 288,317
Kleopfer, Robert, 317 Watters, Robert L . , Jr., 288
K o c h , William F., 210 Wood, Laura J., 288
K u r t z , D a v i d ., 288 Zweig, M a r k H . , 149
Affiliation Index
Bureau of Alcohol, Tobacco, Ministry of Environment, 317

and Firearms, 288 National Bureau
Case Western Reserve University, 70 of Standards, 1,109,171,210,288,317
Colorado State University, 266 National Institutes of Health, 149
Dow Chemical Company, 288 Pennsylvania State University, 288
Ecology & Environment, Inc., 78 Radboud Hospital, 126
Food and Drug Administration, 288 St. John Associates, Inc., 275
Geophysical Laboratory, 275 U.S. A r m y Chemical Research, 224,288
International Atomic Energy Agency, 171 U.S. Environmental Protection Agency, 317
Kaiser Aluminum and Chemical University of Amsterdam, 126
Corporation, 288 University of Georgia, 94
McCormack Associates, 64 University of N e w Hampshire, 194
Yankee Atomic Electric Company, 244
Subject Index
A Acceptable M i n i m u m Detection Amount
Absorption noise sources discussion, 102,103
flicker noise, 119 See also Regulatory limits, process
limiting noises, 120/ specifications
shot noise, 119 Accuracy, definition, 149
328

Author Index
Bauer, C F . , 194 Liggett, Walters., 210

Brossman, Martin W., 317 McCormack, M i k e , 64
Crummett, Warren B., 288 M c K e n n a , Gerard, 317
Currie, L l o y d ., 1,171 Mellor, R . ., 244
Epstein, Michael S., 109 Midkiff, Charles R., Jr., 288
Grant, C . L . , 194 Moss, Thomas H . , 70
Hallowell, S. F., 224 Owens, K . G . , 194
Hanneman, W. William, 288 Parr, Robert M , 171
Hare, P. E . , 275 Reutter, D . J., 224
Harrington, C L . 244 Rogers L B. 94
Horwitz, William
Johnson, Jame
Johnson, Janet ., 266 St. John, P. ., 275
K a h n , Henry, 317 Steigstra, H . , 126
K i n g , Donald, 317 Sturdivan, L a r r y , 288
K i r c h m e r , C l i f f J., 78 Taylor, John K . , 288,317
Kleopfer, Robert, 317 Watters, Robert L . , Jr., 288
K o c h , William F., 210 Wood, Laura J., 288
K u r t z , D a v i d ., 288 Zweig, M a r k H . , 149
Affiliation Index
Bureau of Alcohol, Tobacco, Ministry of Environment, 317

and Firearms, 288 National Bureau
Case Western Reserve University, 70 of Standards, 1,109,171,210,288,317
Colorado State University, 266 National Institutes of Health, 149
Dow Chemical Company, 288 Pennsylvania State University, 288
Ecology & Environment, Inc., 78 Radboud Hospital, 126
Food and Drug Administration, 288 St. John Associates, Inc., 275
Geophysical Laboratory, 275 U.S. A r m y Chemical Research, 224,288
International Atomic Energy Agency, 171 U.S. Environmental Protection Agency, 317
Kaiser Aluminum and Chemical University of Amsterdam, 126
Corporation, 288 University of Georgia, 94
McCormack Associates, 64 University of N e w Hampshire, 194
Yankee Atomic Electric Company, 244
Subject Index
A Acceptable M i n i m u m Detection Amount
Absorption noise sources discussion, 102,103
flicker noise, 119 See also Regulatory limits, process
limiting noises, 120/ specifications
shot noise, 119 Accuracy, definition, 149
328

INDEX 329
A c i d rain, 72,73,76 Blank correction

A l p h a error, See False positive omission, 88
A m i n o acids U n i t e d K i n g d o m policy, 86,91
analysis U n i t e d States policy, 78,86,91
chemiluminescence detection, 283 Blank responses, variability, 78-92
liquid chromatography, 277 Blank variability, propagation
dansyl derivatives, 279 approach, 188
detection limits i n environmental Blood sugar, distribution i n diabetics and
samples, 275-284 nondiabetics, 99,100/
detection using laser-induced Boumans's equation, 118-119
fluorescence, 282
environmental samples, 275,276/
fluorescent derivatives, 279
methods of analysis, history, 278 Calibration curve
post-column chromatograms, 2 8 0 - 2 8 1 / confidence intervals, 204
post-column derivatization, 279 correlation coefficient, 198,198^,199
pre-column derivatization, 279 detection limit estimate 195
underivatized, U V detection
Analyte detection, 9
Analytical blank, 112, 301,302 statistical analysis, 195
Assumptions, testing, 8-9 Calibration error
Atomic absorption spectroscopy, description, 27,291
enhancement, 298 dilution errors, 291,292
Autocorrelation function ( A C F ) , Calibration graphs, nonlinear, 292
mathematical definition, 129 Calibration model
choice, 197
least squares curve, 197
linear, See also Linear model with
Background intercept
relationship of noise components, Calibration standards
Boumans's equation, 118-119 equal variance requirement, 292
variability, 22 preparation, 291,292
Baseline drift Central limit theorem
correction, 143,144,145/ explanation, 21
mathematical model, 143 validity, 21
Beta error Cesium-137 concentrations, log-normal
detection limit, 29 distribution, 267,269/
Beta error, See False negative Chebyshev inequalities, See Gauss
Beta particle concentrations in air, inequality
precipitation, 267,271/ Chebyshev inequality, extension, 61
Beta-gamma systems, background Chemical and Engineering News, 311
frequency distribution, 256,258/ Chemiluminescence detection of amino acids
Bias advantages, 283
blank and calibration factor default description, 283
values, radioactivity monitoring, 34 detection limits, 283
detection limit comparisons, 111 Chromatogram noise, pump cycle vestiges, use
effect on detection limits, 34,35/ of Fourier transform to
introduced by rounding, 28 eliminate, 214,216,218/
Blank Clean A i r A c t , 71,72,76
"blank effect", 187 Clean Water A c t , variability of sample
effective distribution, 23 responses, 87
ideal, 187 Clinical detection, definition, 150
limiting factor, 22-23 Clinical detection limits, 149-170
problems, 21 Clinical laboratory
propagating components, 23 decision analysis, 162
standard deviation, 177,296 signal/noise discrimination, 152
systematic error, 34 test evaluation, 162-167
types, 300-302 Clinical test
variability, 22,188 detection limit, R O C curve, 152
well-known, 84 serum myoglobin concentrations, 152,153

Clinical test evaluation Detection limitContinued

example, detection of marijuana, 168,169/ ion chromatography, 210-222
principles, 162-167 lower and upper, 29,30/
Coefficient of variation, for Sr-90 mathematical formula for variations i n
measurements in soil, 270,271/-273/ background and counting time, 183
Computerized data bases multicomponent, 34
advantages, 317 nuclear measurements, 171-191
discussion, 317-327 Nuclear Regulatory Commission
drawbacks, 318,321 (NRC),172,176/
low-ieve! data, 319-327 optimization, figure of merit ( F O M ) , 34
reporting codes, 318-322 radionuclide, 184
Concentration range, choice, 196 software, 211
Confidence limits uncertainty, 2
nonzero intercept model, 204,205/,206 uses, 210
zero intercept model, 204,205/,206 Determinand, 78
Correlation coefficient, 198,198/, 199 Determination limit, definition, 85
Cost-benefit basis of regulatory decisions, Discrimination limit
$2 million unwritten rule, 7
Covariance among variables, 5
Criterion o f detection, vs limit of definition, 40
detection, 80,83/ description 39,40,41/
Critical level, 4 science applications, 42
Distributions
non-normal, 21
D Poisson, 21
Data evaluation process, fitting of spectral

or chromatographic peaks, 26
Debate cycles
Effects analysis model, explanation, 231,232/
adversary phase, 75
Efficacy, definition, 149-150
risks of naivete, 75
Ensemble
Decision analysis, 162,163
definition, 129
advantages, 7
statistics, 129
See also Decision theory
Equivalent fitting, See Proper pairing
Decision point, 14
Ergodicity, 132
Decision theory
Error
objective function, 7
calibration, 27
See also Decision analysis
components, 23-25
Delaney clause, Federal F o o d , Drug, and
estimates, 23
Cosmetic A c t , 71
internal, precisely known, 25
Detection limit
internal vs external, 23
amino acids in environmental
model, 9
samples, 275-284
propagation vs replication, 23
analytical chemistry
terminology, 23
history, 11
Error variance
See also Historical perspective
basic theory, derivation, 127-133
atomic absorption, discrepancies, 113,114/
function o f integration time, 140,141/
clinical test, R O C curve, 152
comparison of C F A E S and F E S , 114,115/
comparison of F A A S and G F A A S , 113,114/ F
criterion, I U P A C recommendation, 111-112
definitions, 16,17/ Fact manipulation, 72-73
effect o f instrument noise, 113 Failure modes, explanation, 231,232/
example, earthquake, 5,6/,7 False negative
hypothesis testing, 2 example, 3/,4
illustrations for nonscientists, 67-68 reduction to 5%, 80,82/
inequality expression, 36 False positive
International Atomic Energy example, 3/,4
Agency, 172,176/ level of 5%, 80

INDEX 331
Federal Insecticide Fungicide High-temperature gas-cooled power

and Rodenticide A c t reactorContinued
( F I F R A ) debate, 74 comparison o f effluents with those from
Federal Register, 310-311 boiling water reactors and pressured
Feigel water reactors, 266,267/
identification limits, 11,60 radioactivity, 267
spot tests, 11,12/ See also Environmental monitoring
First-order noise, A C F , P S D and Historical perspective, 10-14
record, 137,138/ detection limit terminology, 12/
Flame atomic absorption spectroscopy Feigel, 11,12/
(FAAS) Kaiser, 12, 12/, 13
detection limit, 87 signal/noise discrimination, 151
tin detection limit, 111 Horwitz curve, 308,309/
Flame atomic fluorescence spectroscopy Hypothesis testing, 2,4,9
(FAFS)
cadmium detection, 111
improving detection limit, 112
Flame transmission noise, zinc
Forensic laboratory results,
limitations, 296,297 multivariable, 49
Fourier transform function, definition, 129 Inductively coupled plasma (ICP)
Functional relation, 9 spectrometry
Fundamental detection limits, blank, emission vs wavelength spectral
definition, 110 scan, 302,303/
Fusarium mycotoxins detection limit estimates, 301
analysis, scheme, 227/ discussion, 300
See also Trichothecene mycotoxins instrument comparison, 301
Fusarium trichothecenes, detection interference, 302,303/
i n human blood, 224 Inductively coupled plasma (ICP)-atomic
Fusion blank, 302 emission spectrography, noise, 137,139/
Instrument noise
effect on the detection limit, 113
G
flicker noise limited, 113
Instrumental detection limits,
Gamma analysis, nondestructive use of
mock-up, 298 definition, 110
Gamma spectroscopy Integrated noise
beaker geometry, 259 standard deviation, 146,147/
Ce-144, 260,261/,262 variance, 133-135
Integrated peak and noise, 127,128/
method, 259
Integration limit
releases from pressurized-water nuclear
optimum
power facilities, 259
determination, 140
Ru-106, 260,261/,262
noisy and normalized peak, 140,141/
uncertainty, 260
Integration variance after baseline
Gamma-ray peaks for Hg-203
and Cr-51, 178/ correction, 143
Gamma-ray spectrometry Intercept model, confidence limits,
comparison to radiochemical analysis, 184 nonzero, 204,205/,206
description, 184 Interference
detection limit, pitfalls, 184,185 computer search to determine compounds
Gauss inequality, 21 responsible, discussion, 103-104
Graphite furnace atomic absorption examples, 99-101
spectroscopy ( G F A A S ) , measurement Interlaboratory measurements
of detection limit, 113 arsenic i n horse kidney, 177,178/, 179
atomic absorption spectroscopy, 179
dioxin compounds in human
H fat, 293,294,294/
milk powder, 179
High-temperature gas-cooled power radiochemical neutron activation
reactor analysis, 179
background, 267 variables, 94

Interlaboratory tests, quality control, 188 Limit of detectionContinued

International Atomic Energy Agency, estimation for environmental analytical
systematic error bounds, 187 procedures, 78-92
International Atomic Energy Agency Linear model with intercept
recommendations nonzero intercept, 203
blank and contamination problem, 185 nonzero intercept model, 197
illustration, 186 zero intercept, 201,203
International U n i o n of Pure and A p p l i e d zero intercept model, 197-198
Chemistry ( I U P A C ) , 11 See also Regression analysis
International U n i o n of Pure and Applied Low-level data, 28
Chemistry ( I U P A C ) Nomenclature Lower limit of detection ( L L D ) ,
Document, risk levels, 13 requirements, 173/
Iodine-131 L o w e r limit o f detection
concentration at air sampling determination, 248
stations, 270,271/ N U R E G formulation, 246
detection limit, 270 regulatory definition, 246
low-level analysis R E T S formulation, derivation, 246
frequency distributions fo
recoveries exhibited by
chemists, 256
mass data, historical, 255/ Manganese concentrations in human
method, 251,254
milk, 189,190/
monitoring, 267
Matrix effects
Ion chromatogram
signal masking, 298-299
219,220/
signal suppression, 297,298
adjustment for white noise, 217,218/,220/ Matrix variability
analysis, example, 212,213,215/ arson, 299
correction for pump cycle, 214,215/ explosives, 299
filtered, 219,220/ M a x i m u m contaminant level, 308
removal of baseline, 213,215/ Measurement units, "relative" versus
removal of interfering components, 212 "absolute", 113,114,114/
Ion chromatography Method detection limit ( M D L )
description, 211 as error distribution, 88,89/
software, 219 definition, 307
equation, 88
H P L C , 307
illustration, 307,309/
Kaiser
problems, 307
impact on analytical chemistry, 12/, 13
Methodological detection limits,
terminology for detection limit, 12/, 13
definition, 110
Multichannel identification, 44
Multiple detection decisions
parallel tests, 43
serial tests, 43
Laser-excited atomic fluorescence
Multivariable identification, 44,46/
spectrometry ( L E A F S )
detection limits, 123/
limiting noises, 123/
Laser-excited flame atomic fluorescence
spectrometry ( L E A F S )
description, 121 Neutron activation analysis, analysis of
detection limits, 121 trace elements in human milk, 187,188
Laser-induced fluorescence Neyman-Pearson (frequentist) approach to
detection limits, 282 significance testing, 7
detection of amino acids, 282 Noise
wavelength-tunable lasers, 282 detection-limiting, 116,116/
Likelihood ratio, derivation, 7,8 emission noise sources, 118
L i m i t of detection fast and slowly fluctuating, 129,130/, 132
assumptions, 86 flicker, 118
calculation, 295 1/f component, 137

INDEX 333
NoiseContinued Q
nonstationary, 143
nonwhite, 20 Quality assurance, 231,233/
random, 20 Quantification limit, volatile organic
records, ensemble, 131/, 132 compounds, 308
relationship of components to background,
Boumans's equation, 118-119
shot, 118
signal-carried, definition, 110 Radiological effluent technical
white, 137 specifications ( R E T S ) , 244
Nonwhite noise, 20 Radionuclide detection limits, 184
Nonzero intercept model, confidence Reactor effluent releases, detection, 270
limits, 204,205/,206 Reactor environmental monitoring
Normal random noise, 20 beta concentration, air, 267,268/
Nuclear measurements, detection beta precipitation, 267,271/
limits, 171-191 Cs-137, 267,269/
Nuclear Regulatory Commission discussion, 266-274
estimation of detection limits
programs, synopsis, 189,189
standard deviation of the blank, Reagent blank, 302
definition, 177 Receiver Operating Characteristic ( R O C )
N u l l hypothesis, testing, 18 applications, 162
N u l l hypothesis state, definition, 245 example, myocardial
N u l l signal infarction, 159,160/, 161/
background, 22 clinical test improvement, 31,162,164/
chromatographic baselines, 22 derivation, 154
description, 154,155/, 156
example, 154,155/
frequency distribution, 162,163/
One-tailed test, 320-321 hypothetical, 158/
Operating Characteristic (OC) curve, uses interpretation, 156-157
in medicine and psychology, 31 normal curve, 31,32/
Ordered detection limits, 14,15/ test comparison, 157
Organic carbon content of water, Receiver operating characteristic ( R O C )
procedures for determination, 106/ curve, history, 151
Regression analysis
sums o f squares, 198,199,199/,200/
tables, calculation, 201,202/
Paired comparisons, 80,84,85 Regulatory decisions, cost-benefit basis, 5
85 Regulatory limits
Pattern discrimination limits, see discussion, 5-7
M u l t i v a r i a t e identification lower and upper, 40
Poisson distribution, 21 process specifications
36 American National Standards Institute
Precision, 291 ( A N S I ) , Acceptable M i n i m u m
Probability density function ( P D F ) Detection Amount, 102,103
explanation, 129 general protocol recommended by the
noise fluctuation, 129,130/ Association of Official Analytical
Proper pairing, advantages, 22 Chemists ( A O A C ) , 102
Public trust recommendation, 40
dioxin compounds in human Relative standard deviation (RSD),
fat, 293,294,294/ Poisson, 61
limit of detection, 293 Relative standard deviation (RSD),
Pump cycle factors that increase
dependence on eluent, 216,217,220/ contamination, 95,96/
instability, 216 inexperience, 97
ion chromatogram, 218/ lack of blind quality assurance
power index, 216 samples, 98
variations in strength, 216 participation of more than one person, 97

Relative standard deviation (RSD), Software problemsContinued

factors that increaseContinued quality, 38-39
sample loss, 95,97 thresholds, 38
subjective use of statistics, 97 Specificity
table, 98/ definition, 151
Replication of standards, benefits, 196-197 example, 151
Reporting results Spectrometric system, components, 116,117/
approximation to average, 311-312 Spline, cubic, definition, 213
low-level data for computerized Standard deviation of the blank,
data base,317-327 calculation, 296
negative values, 312,313/ Standards, preparation, 197
Risk, definition, 231 Statistical control
Risk assessment, 229 control charts, 290
failure modes and effects analysis limiting mean, 290
model, 231 stable standard deviation, 290
Risk level, I U P A C recommendation, 13 Sums o f squares
Rounding and truncation, effects, 26 lack of fit, 199,200/,201
Sample manipulation Symbols

quality of data vs steps in definition, 60
procedure, 229,230 list, 10/
risks, 229 Systematic and model error, 20
Sample preparation, 291,292 Systematic error, blank, 34
Selectivity Systematic error bounds, 184
discussion, 99-101
determinations of dioxins, Dow, 101
interferences, 99-101
Self-absorption, 298
Sensitivity
definition, 150,151 t-test, noncentral t, 31
example, 151 Terminology, confusion, 11-13
Sensitivity-specificity pairs, Thresholds, setting, 28
description, 152 Toxic Substance A c t ( T O S C A ) , 73
Signal-carried noise, definition, 110 Trichothecene mycotoxins
Signal-to-background ratio, 118 description, 225
Signal-to-noise ratio, 118 of interest to U.S. A r m y , 226/
Signal-noise discrimination program requirements for analysis, 228
clinical laboratory, 152 See also Fusarium mycotoxins
historical perspectives, 151 Tritium
serum myoglobin analysis, evaluation, 264
concentrations, 152,153/ background frequency distribution for
Signals and noise, example, 151,153/ liquid scintillation unit, 251,253/
Single ion chromatogram measurement protocol
heptafluorobutyl ester of description, 248
scirpentriol, 234,235/ frequency distribution, 248,249/,250/
regression curve, 234,237/ reanalysis, 251,252/
T - 2 toxin, 234,236/
Single species matching, univariable
identification, 44,45/
Societal decisions and actions, 7 U
Sociopolitical perspectives, 51-54
Software Uncertainty, calculation, 33
algorithms, 211 Uncertainty in lower limit of detection
peak identification, 211 background, 248
Software problems empirical, 248
data-dependent algorithms, 38 propagation o f uncertainty, 248
model search, 38 Univariable identification, single species
peak search, 38 matching, 44,45/

INDEX 335
V White noise, See Noise, random
Variance weighted-t technique

advantages, 25
explanation, 25
Variances Zero intercept model
heterogeneity confidence limits, 204,205/,206
Bartlett's Chi-square test, 203 linear, explanation, 195,196
internal, 25 Zero risk concept
weighting, 203-204 benzo(a)pyrene, 306
National Pollution Discharge Elimination
W System ( N P D E S ) , 306-309
Zero-intercept model
Water analysis equation, 201
blank correction omitted, 88,90 theory, 201
variability o f blank responses, Clean See also Regression analysis
Water A c t , 86,87 Z i n c determination by F A A S 121,122/
Production by Car a Aldridge Young

Indexing by Linda R. Ross
Jacket design by Caria L. Clemens
Elements typeset by Hot Type Ltd., Washington, DC

Printed and bound by Maple Press, York, PA


(Lloyd A. Currie (Eds.) ) Detection in Analytical C (Book4You)

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

(Lloyd A. Currie (Eds.) ) Detection in Analytical C (Book4You)

Încărcat de

Drepturi de autor:

Formate disponibile

ACS SYMPOSIUM SERIES 361

Lloyd A. Currie, EDITOR

Developed from a symposium sponsored

American Chemical Society, Washington, DC 1988

In Detection in Analytical Chemistry; Currie, L.;

In Detection in Analytical Chemistry; Currie, L.;

1987 Advisory Board

Alan Elzerman W. H . Norton

John W. Finley James C . Randall

Marye Anne Fox E . Reichmanis

Martin L . Gorbaty C . M . Roland

Roland F. Hirsch W. D. Shults

G . Wayne Ivie Geoffrey K. Smith

Rudolph J. Marcus Douglas B. Walters

In Detection in Analytical Chemistry; Currie, L.;

In Detection in Analytical Chemistry; Currie, L.;

CHEMICAL MEASUREMENTS ARE CHARACTERIZED by three fundamental

In Detection in Analytical Chemistry; Currie, L.;

In Detection in Analytical Chemistry; Currie, L.;

Detection: Overview of Historical, Societal,

Center for Analytical Chemistry, National Bureau of Standards,

T h i s chapter not subject to U . S . copyright

In Detection in Analytical Chemistry; Currie, L.;

The DETECTION LIMIT (Lp) i s one o f t h e most important

In Detection in Analytical Chemistry; Currie, L.;

Background Tremors Earthquake

Ordinary Luggage Explosives

Method B l a n k Toxic Chemical

Volume F l u c t u a t i o n s S t o r a g e Tank Leakage

F i g . 1. H y p o t h e s i s T e s t i n g and D e t e c t i o n L i m i t s . The upper

hypotheses, w i t h the corresponding d e c i s i o n s [D , D ] a t the l e f t . Q A

Two k i n d s o f erroneous d e c i s i o n s may be made: f a l s e p o s i t i v e s

where adequate d e t e c t i o n l i m i t s f o r t h e H 's have c l e a r , p r a c t i c a l

In Detection in Analytical Chemistry; Currie, L.;

s t r a t o s p h e r i c ozone changes, C 0 - i n d u c e d g l o b a l warming), t h e

1. BREADTH: The Scope o f D e t e c t i o n

o c c u r r e d ( a t a g i v e n time and p l a c e ) . To t e s t H , one r e q u i r e s a

t e s t - o r measurement-process (MP), o f t e n a Chemical Measurement

In Detection in Analytical Chemistry; Currie, L.;

radiocarbon concentration [isotope r a t i o ] f o r l i v i n g matter i s

even c o n t i n u o u s v a l u e s ( i n f i n i t e number o f H ' s ) . A Such i s the

and H , and the two k i n d s o f e r r o r , a and p.

1.1 Regulatory Limit

In Detection in Analytical Chemistry; Currie, L.;

and the s i g n a l t o be d e t e c t e d [ H : S-Lj, ] , and a c c e p t a b l e f a l s e

In Detection in Analytical Chemistry; Currie, L.;

the a l g e b r a o f r e g u l a t i o n , a de f a c t o "$2 m i l l i o n u n w r i t t e n r u l e "

In Detection in Analytical Chemistry; Currie, L.;

to that f o r H Q a t the s i g n a l l e v e l i n q u e s t i o n . ] The optimal

the n e t c o s t o f a f a l s e p o s i t i v e and the p r i o r p r o b a b i l i t y o f H Q

d i v i d e d by the p r o d u c t o f the n e t c o s t o f a f a l s e n e g a t i v e and the

1.3 Testing of Assumptions. The detection of erroneous

In Detection in Analytical Chemistry; Currie, L.;

and parameters f o r a) the p h y s i c o c h e m i c a l ( o r e m p i r i c a l ) model and

1.4 Analyte Detection. T h i s i s a p r i m a r y f o c u s f o r t h i s volume,

In Detection in Analytical Chemistry; Currie, L.;

whereas the detection limit refers t o the underlying true

C r i t i c a l Level Detection Limit

ing quantity. A c o r o l l a r y i s t h a t t h e term " d e t e c t i o n l i m i t " i s

The d u a l q u e s t i o n s , "How l i t t l e c a n I d e t e c t ? " , and "Has something

In Detection in Analytical Chemistry; Currie, L.;

o The ad hoc: s e l e c t i n g a r i g i d f o r m u l a , o f t e n based on some

o The s i g n a l / n o i s e : g e n e r a l l y assuming white noise, and