Documente Academic
Documente Profesional
Documente Cultură
To cite this article: A. C. Williams B.D.S., F.D.S., M.C.D.H., D.D.P.H. & C. D. Stephens M.D.S.,
B.D.S., F.D.S., M.Orth. (1992) A Modification to the Incisor Classification of Malocclusion, British
Journal of Orthodontics, 19:2, 127-130, DOI: 10.1179/bjo.19.2.127
Abstract. During the study of two-hundred orthodontic cases under treatment in the General Dental Service,
four clinicians showed only moderate agreement using the British Standard Classification of Incisor
Ma/occlusion. Cohen's Kappa statistic was used to measure the inter-examiner agreement. The 35 cases in
which there was high disagreement between the examiners were scrutinized to determine the source of this
disagreement. Following discussion, revised definitions were produced which included the introduction of a
Class [[-intermediate group. After an interval of 2 months these 35 cases were reclassified using the new
definitions. Increased inter-examiner agreement was found. Four examiners (three of whom were common to
the original study) then used the modified classification to describe the incisal relationship of lOO cases referred
for treatment at Bristol Dental Hospital. The inter-examiner and the intra-examiner agreement were both
found to be good. It is recommended that the Incisor Classification of M a/occlusion be extended to include a
Class !!-intermediate group.
Index words: Classification of Malocclusion, Reproducibility, Reliability.
the reproducibility of the British Standard Incisor TABLE 2 Kappa values for four observers
Classification of Malocc/usion and to introduce using the B.S.I. classification of incisor maloc-
clusion for 200 models
modifications to improve its reliability.
Observer
Materials and Methods B c D
Four experienced clinicians were asked to classify
Observer
independently the incisor relationship of 200 sets of 0·61 0·57 0·64
A
study models chosen at random from those pro- B 0·68 0·61
vided by the Dental Practice Board as part of c 0·53
another study (Shaw et al., 1991). Cohen's Kappa
statistic (Cohen, 1960) was used to measure the Mcan=0·61.
inter-examiner agreement which was classified
according to Land is and Koch ( 1977) (Table 1). The TABLE 3 Incisor cla.!sification of the 35 cases in which fewer than
thirty-five cases in which less than three examiners three examiners agreed
agreed were scrutinized to find the source of the
Number of cases
disagreement. After an interval of several months, Alternatives classified as Class I
these 35 cases were re-examined by the same
examiners using a modified classification. Class Ill 20
As an overall measure of the reliability of the Class 11 I 6
modified classification, the classification was Class 11 2 4
Clss Ill or 11 2 (by different examiners) 5
applied to lOO sets of good quality study models of
patients under treatment at Bristol Dental Hospital Total 35
representing the full range of malocclusion. The
models were examined by four experienced clini-
cians, three of whom were common to the original Confusion also existed as to whether cases in which
study. The same clinicians reclassified the models two incisors were in Class I occlusion and two were
after an interval of 2 weeks to give a measure of the in Class Ill occlusion should be classified as Class I
intra-examiner reliability of the modified classifica- or Class Ill.
tion of incisor relationship. There was disagreement as to whether upright
incisors with an increased overjet should be classi-
fied as Class 11 division 1 or Class 11 division 2 and
Results this was resolved by the introduction of a 'Class 11-
Table 2 shows that agreement between the exa- intermediate category'. This was defined as a maloc-
miners using the British Standard Classification of clusion in which 'The lower incisor edges lie pos-
Incisor Ma/occ/usion to classify 200 sets of study terior to the cingulum plateau of the upper central
models varied from moderate to substantial. incisors. The upper incisors are upright or slightly
Less than three examiners agreed on the classifi- retroclined and the overjet lies between 5 and 7 mm.'
cation of 35 sets of study models in the group. Most When the same examiners used the modified
of the disagreement related to border-line Class I classification to classify the 35 models in which
and Class 11 cases (Table 3) and appeared to arise agreement had previously been slight (Table 4) an
fcom a failure to appreciate that it was the cingulum improved agreement was found (Table 4).
plateau rather than the middle third of the visible The intra-examiner agreement was substantial
palatal surface of the maxillary central incisor when reclassifying the whole sample of lOO models
which is the crucial aspect of the classification. (Table 4).
Discussion
TABLE I Interpretation of kappa values
Classifications are made of biological variation in
Agreement Kappa value order to facilitate communication between clini-
Poor <0·00 cians and as a basis for diagnosis. For a classifica-
Slight 0·00-0·20 tion to be of value it should be simple, objective,
Fair 0·21-0·40 appropriate, reproducible, and reliable. The reliabi-
Moderate 0·41-0·60 lity and reproducibility of the British Standard
Substantial 0·61·0·80
Almost perfect >0·80
Classification has not been tested before despite its
widespread use in clinical practice.
BJO May 1992 Modification of Incisor Classification of Malocclusion 129
TABLE 4 (a) Kappa mlues for four observers and have no pathological basis to them. Because
using the British Standard classification ofincisor malocclusion is an expression of biological varia-
malocclusion for 35 models showing poorest
tion, any classification is subject to the problems
agreement
associated with the morphological variation of the
Observer teeth and the relationship of the dental bases. This is
particularly true of the Class II sub-classifications
B c D (division I and division 2) which are made on the
Observer
basis of overjet and upper incisor angulation. There
A 0·12 0·19 0·34 is a tendency to classify any case with an increased
B 0·32 0·001 overjet as Class II division I when it is quite possible
c 0·0015 to have a Class II division 2 malocclusion where,
although the upper incisors are retroclined, the
Mean=0·16.
overjet is still increased.
(b) Kappa 1•alues for jimr observer~· using the The modified classification of incisor malocclu-
modified classification o.fincisor malocclusionfor sion used in this study introduces a 'Class II-
35 models intermediate' group to overcome this problem
Observer which has specific criteria with specific treatment
implications and is not merely a group that does not
B c D 'fit' into the existing classification. The Class II-
intermediate classification embraces some of those
Observer cases identified as having upright incisors, but an
A 0·58 0·60 0·28
B 0·74 0-41 increased overjet where fixed appliances are
c 0·35 required in order to bodily retract the upper labial
segment. The classification of such cases as Class IJ
Mean=0-49. division I could lead to inappropriate treatment
(c) Kapptl mlues for ill/er-examiner agreement being used particularly by the inexperienced opera-
betwt•enfour examiners using the modified classi- tor.
fication o.f incisor malocclusion for lOO models Gravely and Johnson ( 1974), measured the relia-
bility of Angle's classification of malocclusion
Observer
which also included groups for 'malocclusion
B c D unclassifiablc', 'normal occlusion', and 'Class II-
unccrtain'. Even with the introduction of such
Observer groups, which must add doubt to the value of the
A 0·74 0·75 0·64 classification, the inter-examiner agreement was
B 0·82 0·68
c 0·69 low. The Class 11-uncertain group was defined as
'one in which doubt existed as to whether the case
Mean=0·72 was Class 11 division I or Class 11 division 2'. The
proposed Class II-intermediate group is similar to
the 'Class 11-uncertain' classification, but unlike the
The low inter-examiner agreement between exa- latter there are specific criteria for the former.
miners using the classification was found by chance Gravely and Johnson repeated the reproducibi-
during part of another study. Most of the cases in lity measurement but examiners were asked to place
which there was disagreement between examiners cases that would previously have been classified as
were due to a failure to follow the literal interpreta- Class'II-uncertain in the Class 11 division I group.
tion of the classification accurately. This is partly An improved inter-examiner agreement was found,
due to complacency by experienced clinicians, but but it was still sufficiently low for them to conclude
also due to the subjective descriptions used in the that the 'classification of malocclusions according
BSI classification. For example, Class II division l to Angle's system cannot be carried out with a high
cases are described as those in which the incisors are level of reliability'. The high degree of inter- and
'usually' proclined. Once the problem of inaccurate intra-examiner agreement found in the present
interpretation of the classification had been study suggests that the introduction of a specific
addressed there remained a small but significant Class II intermediate group can improve the reliabi-
group of cases over which the examiners could not lity and the reproducibility of the incisor classifica-
agree on a classification. tion ofmalocclusion in a way which also highlights
The classifications used in orthodontic cases are treatment possibilities.
based on the morphology of occlusal relationships Modification of the incisor classification of
130 A. C. Williams and C. D. Stephens BJO Vol. /9 No. 2