Infection (2015) 43:345352

DOI 10.1007/s15010-015-0771-0

ORIGINAL PAPER

Sepsis instandard care: patients characteristics, effectiveness

ofantimicrobial therapy andpatient outcomea cohort study
FranzRatzinger1 KatharinaEichbichler2 MichaelSchuardt2 IreneTsirkinidou2
DieterMitteregger3 HelmuthHaslacher1 ThomasPerkmann1
KlausG.Schmetterer1 GeorgDoffner4 HeinzBurgmann2

Received: 22 August 2014 / Accepted: 24 March 2015 / Published online: 4 April 2015
Springer-Verlag Berlin Heidelberg 2015

Abstract Results Of the 2384 screened patients, 298 fulfilled

Purpose Fast diagnosis and initiation of appropriate anti-
biotic therapy is pivotal for the survival of sepsis patients.
However, most studies on suspected sepsis patients are con-
ducted in the intensive care unit or in the emergency room
setting, neglecting the standard care setting. This study
evaluated sepsis risk factors, microbiological accurateness
of the initial empiric antimicrobial therapy and its effect on
hospital mortality in standard care patients.
Methods In this prospective observational cohort study,
patients with clinically suspected sepsis meeting two or
more SIRS criteria were screened on standard care wards.
After hospital discharge, occurrence of an infection was
assessed according to standardized criteria, and empirical
antibiotic therapy was evaluated using antibiograms of rec-
ognized pathogens by expert review.
Electronic supplementary material The online version of this
article (doi:10.1007/s15010-015-0771-0) contains supplementary
material, which is available to authorized users.
* Heinz Burgmann
heinz.burgmann@meduniwien.ac.at
1
Division ofMedical andChemical Laboratory Diagnostics,
Department ofLaboratory Medicine, Medical University
ofVienna, Vienna, Austria
2
Division ofInfectious Diseases andTropical Medicine,
Department ofMedicine I, Medical University ofVienna,
Vienna, Austria
3 mellitus, chronic lung diseases, or chronic kidney disease
Division ofClinical Microbiology, Department ofLaboratory
Medicine, Medical University ofVienna, Vienna, Austria
4 identification of patients at risk of severe infection. Rapid
Section forArtificial Intelligence, Center forMedical
Statistics, Informatics andIntelligent Systems, Medical
University ofVienna, Vienna, Austria
two or more SIRS criteria. Among these were 28.2%
SIRS patients without infection, 46.3% non-bacteremic/
fungemic sepsis patients and 25.5% bacteremic/fungemic
sepsis patients. Occurrence of a malignant disease and
chills were associated with a higher risk of patients having
bacteremic/fungemic sepsis, whereas other described risk
factors remained insignificant. In total, 91.1% of suspected
sepsis patients received empirical antimicrobial therapy,
but 41.1% of bacteremic sepsis patients received inappro-
priate therapy. Non-surviving bacteremic sepsis patients
received a higher proportion of inappropriate therapy than
those who survived (p=0.022).
Conclusions A significant proportion of bacteremic sepsis
patients receive inappropriate empiric antimicrobial ther-
apy. Our results indicate that rapid availability of micro-
biological results is vital, since inappropriate antimicrobial
therapy tended to increase the hospital mortality of sepsis
patients.
Keywords Sepsis Standard care Blood culture
Antibiogram Antimicrobial resistance pattern Mortality
Introduction
Sepsis is a frequent and severe condition, with a high
mortality rate [1, 2]. According to the literature, risk fac-
tors for developing sepsis are higher age, implanted medi-
cal devices or chronic medical conditions such as diabetes
[37]. These risk factors should be kept in mind for faster
pathogen identification as well as early initiation of accu-
rate empiric antimicrobial therapy is of pivotal importance

13
346
for the survival of sepsis patients [8, 9]. Due to emerging
occurrence of multi-drug resistant (MDR) pathogens, the
efficacy of the empiric antimicrobial therapy has to be care-
fully evaluated with respect to a pathogens antibiogram.
Most diagnostic studies evaluating the microbiological
appropriateness of the initial antimicrobial therapy in sep-
sis patients were performed in the intensive care unit or
in the emergency setting, neglecting standard care wards
[10]. Yet it is on standard care wards that a significant pro-
portion of patients at risk of severe infections are treated
[11]. Hence, accurate data are required regarding the pre-
test probabilityfor a positive blood culture result or the
occurrence of an infectionof standard care patients ful-
filling two or more SIRS criteria, their risk factors and their
clinical presentation.
The current study aims to address these issues in the
standard care setting and further evaluates the initial
empiric antimicrobial therapy with respect to microbiologi-
cal susceptibility as well as its effect on hospital mortality.
Materials andmethods
Study design andstudy endpoints
The current study was designed as a prospective observa-
tional cohort study conducted between July 2011 and April
2012 at Vienna General Hospital, a 2112-bed university-
affiliated tertiary care center. Patients on 14 medical and
13 surgical wards with suspected sepsis were screened for
fulfilment of two or more SIRS criteria as outlined by the
ACCP/SCCM conference [12]. As described earlier [13],
all adult inpatients on medical or surgical standard care
wards undergoing blood culture collections were invited
to participate in this study. Exclusion criteria were: age
less than 18years, inability to give consent, HIV-positive
patients, and post-surgical patients (within 72h prior to
the inoculation of blood cultures). Neutropenia related to
therapy was not considered as a valid SIRS criterion. Bac-
teraemia was defined by the detection of a bacterial species
considered a pathogen in blood culture analysis, or by pol-
ymerase chain reaction (PCR). Coagulase-negative staphy-
lococci (CNS) were considered as invasive pathogens only
when detected in blood specimens taken on two sepa-
rate occasions. Blood culture contaminants were defined
according to Hall and Lyman [14]. Occurrence of infec-
tion was assessed following a patients hospital discharge
by applying the hospital-acquired infection definition cri-
teria of the European Centre of Disease Control (ECDC),
which were provided for point prevalence studies [15].
These criteria comprise clinical data as well as laboratory,
microbiological and radiological results. Criteria for clas-
sifying SIRS patients without infections could not be found
13
F. Ratzinger etal.
in the literature. Susceptibility of blood culture pathogens
to empiric antimicrobial therapy was assessed, according to
antibiograms by a board-certified consultant for infectious
diseases as well as a clinical microbiologist of the Depart-
ment of Clinical Microbiology of Vienna General Hospital.
Ethical considerations anddata collection
The study was approved by the local ethics committee of
the Medical University of Vienna (EC-No. 518/2011) and
was conducted in accordance with the Declaration of Hel-
sinki (including current revisions) and the Good Clinical
Practice guidelines of the European Commission.
Clinical information was recorded at the time of patients
enrolment to the study, and was completed after hospital
discharge using the individual medical chart. Blood speci-
mens were cultured in a set of FA Plus (aerobic) and FN
Plus (anaerobic) blood culture bottles in the BacT/ALERT
3D automated blood culture system (bioMrieux, Marcy
lEtoile, France). Isolates were identified by matrix-assisted
laser desorption ionisation (MALDI) time of flight (TOF)
mass spectroscopy (MS) using microflex LT together with
the Biotyper database (Bruker Daltonik GmbH, Bremen,
Germany), VITEK 2 (bioMrieux), or in the case of Strep-
tococcus pneumoniae by additional tests (colony morphol-
ogy on blood agar, optochin disc test, bile salt solubil-
ity test). In some cases, microbial DNA was detected by
SeptiFast MGRADE tests (SeptiFast MGRADE tests, Roche
Diagnostics GmbH, Mannheim, Germany) or SepsiTest
(Sepsitest; Molzym GmbH & Co. KG, Bremen, Germany).
Antimicrobial susceptibility testing was performed accord-
ing to the current recommendations of the European com-
mittee on antimicrobial susceptibility testing (EUCAST,
http://www.eucast.org/). White blood cell counts (WBC)
were measured on a Stromatolyser-4DS (Sysmex, Norder-
stedt, Germany; LLOQ: not provided). All laboratory work
was performed in compliance with ISO 9001:2008 and EN
ISO 15189:2008 in a certified medical laboratory.
Statistical analysis
Data were analysed using the R software version 3.0.2 [16].
Categorical data are presented as counts and percentages,
continuous variables as means and standard deviations. For
univariate analysis, a generalized Fishers exact test [17]
and analysis of variance (ANOVA) according to Rand Wil-
cox were applied [18]. This method used trimmed means
(0.2) as well as bootstrapping (n =2000) for robust cal-
culation of group differences [19]. For post hoc testing, a
robust Students t test according to Wilcoxon with adjusted
p-values was applied. Statistical significance was pre-
defined as a p-value of less than 0.05 (two-sided for non-
directional, one-sided for directional hypotheses). Where
Sepsis in standard care: Patients characteristics
Fig.1Patient recruitment process. *including cases with coagulase-
negative staphylococci in a single blood culture
appropriate, the BonferroniHolm method was used to cor-
rect an accumulation of the -error probability related to
multiple testing.
Results
Patient characteristics andclinical presentation
Between July 2011 and April 2012, 2384 patients with
suspected severe infection were screened. Of these, 444
patients were found to fulfil two or more SIRS criteria
and of these, 298 patients were included in this study (see:
Fig.1). Among the included patients, 249 patients (83.6%)
were treated on medical wards and 49 patients (16.4%)
on surgical standard care wards. 214 participants (71.8%)
suffered from sepsis and 84 patients (28.2%) from SIRS
without any evidence of infection. Most SIRS cases were
related to haematological (n=25, 29.8%) or solid organ
malignant disease (n =13, 15.5%) and cardiomyopathy
(n=9, 10.7%). Furthermore, cases of embolism or bleed-
ing (n=8, 9.5%), auto-immune mediated diseases (n=8,
9.5%), acute organ rejection reaction (n=5, 6.0%), acute
pancreatitis (n =3, 3.6%) as well as other pathologies
(n=13, 15.4%) were recorded.
Most sepsis cases were related to respiratory tract infec-
tion (n =65, 30.4%), gastrointestinal system infection
(n=41, 19.2%) and urinary tract infection (n=32, 15.0%).
Among the 214 sepsis patients, 74 (34.5%) presented with
bacteremic sepsis and two patients presented with fungemic
sepsis (0.9%). In bacteremic/fungemic patients, in most
cases, the infection focus could not be determined (n=23,
30.3%). Major causes of blood-born microorganisms were
central vascular catheter-related infection (n=13, 17.1%)
347
and gastrointestinal tract infections (n =10, 13.2%). The
distribution of infections according to the ECDC classifica-
tion is shown in supplementary Table1.
According to clinical presentation, there was no differ-
ence in parameters used for the SIRS definition between
patients with SIRS, non-bacteremic/fungemic or bactere-
mic/fungemic sepsis. The best discriminatory SIRS param-
eter was heart rate with an observable, butafter apply-
ing the BonferroniHolm correctionnon-significant
difference between the outcome groups (p =0.035, see:
Table 1). SIRS patients without any evidence of infection
had a mean heart rate of 96.1 (11.6), non-bacteremic/
fungemic sepsis patients 96.9 (18.3) and bacteremic/
fungemic patients 102.0 (18.6). Table1 presents numeri-
cally scaled parameters and Table2 nominal or ordinal data
including patients characteristics, clinical parameters and
outcome data. Also, no differences between the outcome
groups were found in epidemiological parameters, includ-
ing patients age, sex, body mass index (BMI) or type of
hospital admission.
Risk factor analysis andpatient outcome
We evaluated several risk factors for the occurrence of
sepsis described in the literature (see: Table2). Signifi-
cant differences between the groups were found in the
occurrence of malignant disease and chills. In the SIRS
group, 54.8% suffered from a current malignant disease,
compared to 28.3% in the non-bacteremic/fungemic sep-
sis group and 51.3% in the bacteremic/fungemic sepsis
group (p<0.001). The occurrence of chills significantly
increased from the SIRS group (44.0%) and the non-bac-
teremic/fungemic sepsis group (44.2%) to the bactere-
mic/fungemic sepsis group (67.1%, p =0.003). Interest-
ingly other risk factors for developing a severe infection
described in the literature, including the rate of diabetes
mellitus, chronic haemodialysis, indwelling catheter, dysu-
ria or diarrhoea, remained insignificantly distributed in uni-
variate analysis. The proportion of antimicrobial treatment
prior to blood culture and empiric antimicrobial therapy
initiation was unevenly distributed within the study popula-
tion. However, after the BonferroniHolm correction was
applied, neither of these parameters showed a significant
difference between the indicated groups (both: p=0.009,
see Table2). Also, outcome parameters such as the admis-
sion rate to ICU (within seven days after study inclusion,
p=0.310), or in-hospital mortality (p=0.753) were dis-
tributed similarly between the outcome classes.
Pathogen spectrum andempiric antimicrobial therapy
Of a total of 298 patients, bloodstream pathogens were
detected in 76 (25.5%), 74 patients had bacteraemia and
13
348 F. Ratzinger etal.

Table1Numerically scaled clinical and laboratory parameters

Descriptive statistics ANOVA
All SIRS Non-bacteremic/ Bacteremic/ Overalla p-valueb p-valuec p-valued
fungemic sepsis fungemic sepsis

N 298 (100%) 84 (28.2%) 138 (46.3%) 76 (25.5%)

Age 56.417.0 53.516.3 57.217.5 58.116.7 0.212 0.108 0.103 0.889
BMIe 25.55.9 26.15.4 25.45.9 25.16.4 0.296 0.299 0.141 0.437
LOSf 22.622.1 24.123.3 21.723.8 22.716.9 0.065 0.056 0.910 0.042
BT Cg 38.31.0 38.11.1 38.30.9 38.61.1 0.059 0.546 0.033 0.049
HR/minh 97.916.9 96.111.6 96.918.3 102.018.6 0.035 0.679 0.018 0.045
RR/mini 21. 15.9 20.15.2 21.46.3 21.55.8 0.461 0.288 0.222 0.765
WBC G/lj 14.232.8 20.760.6 11.97.0 11.27.3 0.466 0.413 0.928 0.261

* Statistically significant after BonferroniHolm correction

a
Robust ANOVA analysis of variance
b
Post-hoc test between SIRS and non-bacteremic/fungemic sepsis
c
Post-hoc test between SIRS and bacteremic/fungemic sepsis
d
Post-hoc test between non-bacteremic/fungemic sepsis and bacteremic/fungemic sepsis; p-values are adjusted for their family-wise error rate
(FWR) with p=0.017 as the critical value
e
Body mass index
f
Lengh of hospital stay
g
Body temperature C
h
Heart rate
i
Respiration rate
j
White blood cells G/l

Table2Ordinal scaled clinical parameters

Parameter All SIRS Non-bacteremic/ Bacteremic/ p-value
fungemic sepsis fungemic sepsis

N 298 (100%) 84 (28.2%) 138 (46.3%) 76 (25. 5%)

Risk factors
Male 169 (56.7%) 50 (59.5%) 82 (59.4%) 37 (48.7%) 0.271
Surgical ward 49 (16.4%) 11 (13.1%) 20 (14.5%) 18 (23.7%) 0.150
COPD 41 (13.7%) 10 (11.9%) 24 (17.4%) 7 (9.2%) 0.247
Diabetes mellitus 46 (15.4%) 9 (10.7%) 25 (18.1%) 12 (15.8%) 0.3
Chronic haemodialysis 18 (6.0%) 7 (8.3%) 6 (4.3%) 5 (6.6%) 0.446
Indwelling catheterb 83 (27.9%) 19 (23.4%) 33 (23.9%) 31 (40.8%) 0.017
Current malignant disease 124 (41.6%) 46 (54.8%) 39 (28.3%) 39 (51.3%) <0.001a
Consequence factors
Chills 149 (50.0%) 37 (44.0%) 61 (44.2%) 51 (67.1%) 0.003a
Dysuria 30 (10.4%) 2 (2.4%) 18 (13.0%) 10 (13.2%) 0.012
Diarrhoeac 67 (22.5%) 15 (17.9%) 35 (25.4%) 17 (22.4%) 0.442
Empiric antimicrobial therapy 261 (87.6%) 66 (78.6%) 128 (92.8%) 67 (88.2%) 0.009
Antimicrobials before blood culture 56 (18.7%) 21 (25.0%) 29 (21.0%) 6 (7.9%) 0.009
ICU stayd 13 (4.4%) 2 (2.4%) 9 (6.5%) 2 (2.6%) 0.310
In-hospital mortality 34 (11.4%) 10 (11.9%) 14 (10.1%) 10 (13.2%) 0.753
a
Statistically significant after BonferroniHolm correction
b
Including: central venous catheter (n=32), urine catheter (n=26), port-a-cath catheter (n=11), permcath catheter (n=2), central venous
catheter with urine catheter (n=8), port-a-cath catheter with urine catheter (n=3), central venous catheter with port-a-cath catheter (n=1)
c
Diarrhoea was assessed according to patients self-assessment
d
ICU stay within seven days after study inclusion

13
Sepsis in standard care: Patients characteristics 349

Table3Distribution of Genus n % Species/resistance mechanism n

pathogen isolates recovered
from blood culture Gram-positive isolates
Staphylococcus 21 23.6 S. epidermidis 5
S. haemolyticus 2
S.aureus 13
S.aureus/MRSA 1
Enterococcus 12 13.5 E. faecium 7
E. faecalis 4
Not specified Enterococcus 1
Streptoccocus 9 10.0 Viridans Streptococcus StrepSStrStreptococcus 5
S. pneumoniae 2
Beta-hemolytic S. 2
Othersa 3 3.4 3
Gram-negative isolates
Escherichia 21 23.6 E. coli 16
E. coli ESBL 4
E. coli AmpC-BL 1
Klebsiella 10 11.2 K. pneumoniae 7
K. pneumoniae AmpC-BL ohne AmpC-BL 1
K. oxytoca 1
R. ornithinolytica 1
Citrobacter 3 3.4 C. braakii, C. freundii 3
Serratia 1 1.1 S. marcescens 1
Proteus 1 1.1 P. mirabilis 1
Acinetobacter Pseudomonas 4 4.5 A. baumannii 2
P. aeruginosa 2
Othersb 2 2.3 2
Fungi
Candida 2 2.3 C. albicans 2
N 89 89

89 isolates in 76 bacteremic/fungemic sepsis patients; in ten patients more than one pathogen was isolated
a
Bacilluscereus, Corynebacterium urealyticum, Clostridium clostridioforme
b
Bacteroides fragilis or not specified Gram-negative rods

2 had fungaemia. In ten patients, more than one pathogen
was isolated. Table3 presents an overview on the distribu-
tion of pathogens. Of 45 Gram-positive isolates (50.6%),
21 were staphylococci, 12 were enterococci and nine were
streptococcus isolates. Within the staphylococci, seven
were CNS and 14 were Staphylococcus aureus isolates
including one MRSA (Methicillin-resistant Staphylococcus
aureus). No case of a vancomycin resistant enterococcus
(VRE) was found. Of 42 Gram-negative isolates (47.2%),
36 were enterobacteria (including 21 Escherichia coli, ten
Klebsiella sp., and one AmpC-positive Klebsiella pneumo-
niae). Four Escherichia coli isolates presented with phe-
notypic evidence of the presence of an extended spectrum
-lactamase (ESBL) and one presented with evidence of
an AmpC -lactamase. Two isolates of Candida albicans
(2.2%) were sensitive to all antifungal substances tested.
Empirical antimicrobial therapy was administered to
195 patients with sepsis (91.1%) and 66 SIRS patients
(78.6%). Of those patients, 180 received -lactam anti-
microbials administered alone or in combination with
other antimicrobials. In total, 60 patients received a ther-
apy regime with an antimicrobial drug combination. The
most prevalent combination contained a -lactam drug
with an antifungal drug (n =15) or with metronidazole
(n =9). Appropriateness of initial antimicrobial therapy
with respect to pathogen susceptibility was evaluated in
73 episodes of bacteraemia (see: supplementary Fig.1).
One bacteremic sepsis patient was excluded from analysis,
because the blood-stream pathogen (Klebsiella pneumo-
niae) was only detected by real-time PCR and therefore no
antibiogram was available. Of these patients, 54 (74.0%)
received appropriate antimicrobial therapy and 19 (26.0%)

13
350
inappropriate empiric antimicrobial therapy, in terms of
wild-type pathogen susceptibility. In the latter group, ten
patients (13.7%) received no or undocumented empiric
antimicrobial therapy and nine patients (12.3%) received
antimicrobial therapy that proved to be ineffective against
the detected pathogen due to wild-type resistance. In a fur-
ther twelve patients (16.4%), the antimicrobial therapy
was ineffective against the detected pathogen, apparently
because of acquired resistance mechanisms. One patient
with Enterococcus faecium in the blood culture responded
clinically to empirical therapy with ampicillin-sulbactam,
even though this species is resistant. Among the bacteremic
patients with inappropriate empirical antimicrobial therapy,
Staphylococcus aureus (n =8, including one MRSA iso-
late) and Escherichia coli (n=7, including two ESBL iso-
lates and one AmpC--lactamase positive isolate) was the
most frequently detected pathogen. Six of the nine bacte-
remic sepsis patients (66.7%) who died during their hospi-
tal stay received inappropriate antimicrobial therapy, while
this proportion was significantly lower among surviving
patients (24 out of 64 patients deceased, 37.5%, p=0.022, while the rate of Gram-negative isolates is increasing [29].
one-sided).
Among non-bacteremic sepsis patients, the causative
pathogen was isolated from various body compartments
in 49 cases. In terms of antimicrobial effectiveness with
respect to wild-type pathogen susceptibility, 15 patients
(30.6%) received inappropriate empiric antimicrobial ther-
apy. Of these patients, two patients received no or undocu-
mented empiric antimicrobial therapy and a further three
received drugs with intermediate effectiveness in antimicro-
bial susceptibility testing. Due to the presence of acquired
resistance mechanisms, the empirical antimicrobial therapy
was ineffective in a further two patients. However, antimi-
crobial susceptibility was not tested for patients specific
antimicrobial therapy in 13 cases (26.5%).
Discussion
In the current study, epidemiological, clinical and labora-
tory data of standard care patients at risk of severe infec-
tions were evaluated. Among the 298 participants, 28.2%
suffered from SIRS without infection, 46.3% from non-
bacteremic/fungemic sepsis and 25.5% from bacteremic/
fungemic sepsis. Most clinical parameters were not use-
ful for categorization into the three outcome groups. As
an exception, bacteremic/fungemic sepsis patients suf-
fered more frequently from chills. Further, patients with a
malignant disease, more frequently suffered from SIRS or
bacteremic/fungemic sepsis, but less frequently from non-
bacteremic/fungemic sepsis. This finding could be based
on an immunosuppression and therefore reduced ability
to control a local infection [2022]. Further, a malignant
13
F. Ratzinger etal.
disease itself or antineoplastic chemotherapy are causa-
tive factors for developing a SIRS syndrome without any
infection, e.g., neutropenic fever [23]. As a SIRS definition
parameter, the heart rate was slightly but insignificantly
higher in patients with bacteremic/fungemic sepsis than in
SIRS patients or patients with non-bacteremic/fungemic
sepsis. In the literature, heart rate appears to have the best
discriminatory potency within the SIRS criteria parameter
[24, 25], which might be related to the level of endotoxine-
mia in patients with sepsis [2628].
Microbiological evaluation
Among the patients (n =76) with bacteremic/fungemic
sepsis, Escherichia coli (n =21), and Staphylococcus
aureus (n=14) were the most frequently detected patho-
gens. While, five multi-drug resistant (MDR) Escherichia
coli isolates (23.8%, ESBL: n =4, AmpC-BL: n =1)
were detected, only one MRSA isolate was found. Gener-
ally, the rate of Gram-positive MDR isolates is decreasing,
Following current sepsis guidelines [30], broad-spectrum
antimicrobials were administered to most patients. How-
ever, 41.1% of sepsis patients with a true-positive blood
culture result received an inappropriate empiric antimicro-
bial therapy regime.
Importantly, a higher proportion of non-surviving bac-
teremic sepsis patients received inappropriate empiric
antimicrobial therapy than those who survived (66.7 vs.
37.5%, p =0.022, one-sided). The association between
inappropriate empiric antibiotic therapy and increased hos-
pital mortality is shown for several patient cohorts, includ-
ing ICU patients [31, 32], patients in the emergency room
[33] or in solid organ recipients [34]. However, hardly any
data are available for standard care setting patients. Gen-
erally, the high percentage of patients receiving inappro-
priate empiric antimicrobial therapy highlights the need
for fast pathogen detection results as well as antibiogram
availability. In a retrospective cohort study with 84 bacte-
remic patients performed at an emergency department, the
initial antimicrobial therapy would have had to be changed
in 21% of cases due to its microbiological unsuitability
for the wild-type sensitivity of the respective isolates [35].
This proportion is in the range of our study, with 26.7% of
initial empiric antimicrobial therapy proving inappropriate
for wild-type sensitivity. In recent years, several technically
advanced methods, including MALDI-TOF MS or molec-
ular methods, have extended the possibilities and timeli-
ness of pathogen detection [36]. Routine application of
MALDI-TOF and molecular methods for faster detection
of antimicrobial resistance patterns of detected pathogens
is currently under investigation [3638]. Our results would
suggest that their broad availability for routine diagnostics
Sepsis in standard care: Patients characteristics
is to be recommended. Moreover, in a significant propor-
tion of non-bacteremic/fungemic sepsis patients (26.5%),
the susceptibility of the empirically administered antimi-
crobial substance was not covered by the antimicrobial
susceptibility test results. In this regard, a harmonization
between applied antimicrobial prescription practice, anti-
microbial stewardship programs and the patterns tested in
antimicrobial resistance tests would be desirable.
Limitations
Several study limitations must be disclosed in this observa-
tional cohort study. Due to the monocentric design of the
study, these results cannot be generalized without restric-
tions. Since the blood culture request was used as the pri-
mary selection criterion for screening of potential study
participants, a selection bias cannot be ruled out. To focus
on a relevant patient cohort, solely patients with two or
more SIRS criteria were included in the study. This restric-
tion resulted in a limited sample size. Finally, antimicrobial
treatment prior to the taking of blood cultures was less fre-
quently found in bacteremic/fungemic sepsis patients. Since
starting antimicrobial treatment before having taken patient
samples for microbiological diagnostics is a well-known yet
unsolved and apparently unavoidable predicament in clinical
routine, it was not assessed as a major limitation. In addition,
the blood culture flasks used contain resin particles, which
decreased the concentration of or completely absorbed anti-
microbial substances present in patients blood [39].
Conclusion
A significant proportion of patients receive inappropriate
empiric antimicrobial therapy according to microbiologi-
cal susceptibility criteria. Therefore, faster availability of
microbiological results including antimicrobial resistance
patterns, especially of blood culture analysis, is warranted,
since appropriate antimicrobial therapy is important for the
patients survival and convalescence.
Acknowledgments We wish to thank Dr. Wolfgang Barousch for
providing microbiological test results and Dr. John Heath for attentive
editing of our English manuscript.
Conflict of interestNone.
References
1. Martin GS, Mannino DM, Eaton S, Moss M. The epidemiology
of sepsis in the United States from 1979 through 2000. N Engl J
Med. 2003;348:154654. doi:10.1056/NEJMoa022139.
351
2. Laupland KB, Svenson LW, Gregson DB, Church DL. Long-term
mortality associated with community-onset bloodstream infection.
Infection. 2011;39:40510. doi:10.1007/s15010-011-0149-x.
3. Yoshida T, Tsushima K, Tsuchiya A, Nishikawa N, Shirahata K,
Kaneko K, etal. Risk factors for hospital-acquired bacteremia.
Intern med (Tokyo, Japan). 2005;44:115762.
4. Wang HE, Shapiro NI, Griffin R, Safford MM, Judd S, Howard
G. Chronic medical conditions and risk of sepsis. PLoS one.
2012;7:e48307. doi:10.1371/journal.pone.0048307.
5. Esper AM, Moss M, Lewis CA, Nisbet R, Mannino DM, Mar-
tin GS. The role of infection and comorbidity: factors that influ-
ence disparities in sepsis. Crit Care Med. 2006;34:257682.
doi:10.1097/01.ccm.0000239114.50519.0e.
6. Wang HE, Gamboa C, Warnock DG, Muntner P. Chronic kid-
ney disease and risk of death from infection. Am J Nephrol.
2011;34:3306. doi:10.1159/000330673.
7. Bos MM, Smeets LS, Dumay I, de Jonge E. Bloodstream
infections in patients with or without cancer in a large com-
munity hospital. Infection. 2013;41:94958. doi:10.1007/
s15010-013-0468-1.
8. Ferrer R, Martin-Loeches I, Phillips G, Osborn TM, Townsend
S, Dellinger RP, etal. Empiric antibiotic treatment reduces mor-
tality in severe sepsis and septic shock from the first hour: results
from a guideline-based performance improvement program. Crit
Care Med. 2014;. doi:10.1097/ccm.0000000000000330.
9. Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S,
etal. Duration of hypotension before initiation of effective anti-
microbial therapy is the critical determinant of survival in human
septic shock. Crit Care Med. 2006;34:158996. doi:10.1097/01.
ccm.0000217961.75225.e9.
10. Mearelli F, Orso D, Fiotti N, Altamura N, Breglia A, De Nardo
M etal. Sepsis outside intensive care unit: the other side of the
coin. Infection. 2014:111. doi:10.1007/s15010-014-0673-6.
11. Brunkhorst FM. Epidemiology, economy and practiceresults of
the German study on prevalence by the competence network sepsis
(SepNet). Anasthesiologie Intensivmedizin Notfallmedizin Schmer-
ztherapie : AINS. 2006;41:434. doi:10.1055/s-2005-921227.
12. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus
WA, etal. Definitions for sepsis and organ failure and guide-
lines for the use of innovative therapies in sepsis. The ACCP/
SCCM Consensus Conference Committee. American College
of Chest Physicians/Society of Critical Care Medicine. Chest.
1992;101:164455.
13. Ratzinger F, Schuardt M, Eichbichler K, Tsirkinidou I, Bauer M,
Haslacher H, etal. Utility of sepsis biomarkers and the infection
probability score to discriminate sepsis and systemic inflamma-
tory response syndrome in standard care patients. PLoS one.
2013;8:e82946. doi:10.1371/journal.pone.0082946.
14. Hall KK, Lyman JA. Updated review of blood culture contami-
nation. Clin Microbiol Rev. 2006;19:788802. doi:10.1128/
CMR.00062-05.
15. ECDC. 2006 (Protocol version 4.3.).
16. R Development Core Team, R: A language and environment for
statistical computing. R Foundation for Statistical Computing,
Vienna, Austria. ISBN 3-900051-07-0. 2005.
17. Luke J. West RKSH. exact tests for two-way contingency tables
with structural zer4os. J Stat Softw. 2008;28:119.
18. Wilcox RR, Schnbrodt FD. The {WRS} package for robust sta-
tistics in {R}. 2009.
19. Carpenter J, Bithell J. Bootstrap confidence intervals:
when, which, what? A practical guide for medical statisti-
cians. Stat Med. 2000;19:114164. doi:10.1002/(SICI)1097-
0258(20000515)19:9<1141:AID-SIM479>3.0.CO;2-F.
20. Vento S, Cainelli F. Infections in patients with cancer undergo-
ing chemotherapy: aetiology, prevention, and treatment. Lancet
Oncol. 2003;4:595604. doi:10.1016/s1470-2045(03)01218-x.
13
352
21. Bohlius J, Herbst C, Reiser M, Schwarzer G, Engert A. Gran-
ulopoiesis-stimulating factors to prevent adverse effects in the
treatment of malignant lymphoma. Cochrane Database Syst Rev.
2008;4:235. doi:10.1002/14651858.CD003189.pub4.
22. Bow EJ. Infection in Neutropenic Patients with Cancer. Crit Care
Clin. 2013;29(3):41141. doi:10.1016/j.ccc.2013.03.002.
23. Bow EJ. Infection risk and cancer chemotherapy: the impact of
the chemotherapeutic regimen in patients with lymphoma and
solid tissue malignancies. J antimicrob chemother. 1998;41:15.
24. Jones GR, Lowes JA. The systemic inflammatory response syn-
drome as a predictor of bacteraemia and outcome from sepsis.
QJM mon j Assoc Physicians. 1996;89:51522.
25. Mato A, Fuchs BD, Heitjan DF, Mick R, Halpern SD, Shah PD,
etal. Utility of the systemic inflammatory response syndrome
(SIRS) criteria in predicting the onset of septic shock in hospital-
ized patients with hematologic malignancies. Cancer Biol Ther.
2009;8:1095100.
26. Scheff JD, Mavroudis PD, Calvano SE, Lowry SF, Androula-
kis IP. Modeling autonomic regulation of cardiac function and
heart rate variability in human endotoxemia. Physiol Genomics.
2011;43:95164. doi:10.1152/physiolgenomics.00040.2011.
27. Jellema WT, Veerman DP, De Winter RJ, Wesseling KH, Van
Deventer SJ, Hack CE, etal. In vivo interaction of endotoxin
and recombinant bactericidal/permeability-increasing protein
(rBPI23): hemodynamic effects in a human endotoxemia model.
J lab clin med. 2002;140:22835.
28. Mignon F, Piagnerelli M, Van Nuffelen M, Vincent JL. Effect
of empiric antibiotic treatment on plasma endotoxin activ-
ity in septic patients. Infection. 2014;42:5218. doi:10.1007/
s15010-014-0586-4.
29. Giske CG, Monnet DL, Cars O, Carmeli Y. Clinical and eco-
nomic impact of common multidrug-resistant gram-negative
bacilli. Antimicrob Agents Chemother. 2008;52:81321.
doi:10.1128/aac.01169-07.
30. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal
SM, etal. Surviving sepsis campaign: international guidelines for
management of severe sepsis and septic shock: 2012. Crit Care
Med. 2013;41:580637. doi:10.1097/CCM.0b013e31827e83af.
13
F. Ratzinger etal.
31. Kollef MH, Sherman G, Ward S, Fraser VJ. Inadequate antimi-
crobial treatment of infections: a risk factor for hospital mortal-
ity among critically ill patients. Chest. 1999;115:46274.
32. Kollef MH. Inadequate antimicrobial treatment: an important
determinant of outcome for hospitalized patients. Clin Infect Dis.
2000;31:S1318. doi:10.1086/314079.
33. Chen HC, Lin WL, Lin CC, Hsieh WH, Hsieh CH, Wu MH,
etal. Outcome of inadequate empirical antibiotic therapy in
emergency department patients with community-onset blood-
stream infections. J antimicrob chemother. 2013;68:94753.
doi:10.1093/jac/dks475.
34. Lupei MI, Mann HJ, Beilman GJ, Oancea C, Chipman JG. Inad-
equate antibiotic therapy in solid organ transplant recipients is
associated with a higher mortality rate. Surg infect. 2010;11:33
9. doi:10.1089/sur.2008.076.
35. Stoneking LR, Patanwala AE, Winkler JP, Fiorello AB, Lee ES,
Olson DP, etal. Would earlier microbe identification alter anti-
biotic therapy in bacteremic emergency department patients? J
emerg med. 2013;44:18. doi:10.1016/j.jemermed.2012.02.036.
36. van Belkum A, Durand G, Peyret M, Chatellier S, Zam-
bardi G, Schrenzel J, etal. Rapid clinical bacteriology and its
future impact. Ann lab med. 2013;33:1427. doi:10.3343/
alm.2013.33.1.14.
37. Sparbier K, Lange C, Jung J, Wieser A, Schubert S, Kostrzewa
M. MALDI biotyper-based rapid resistance detection by stable-
isotope labeling. J Clin Microbiol. 2013;51:37418. doi:10.1128/
jcm.01536-13.
38. Sparbier K, Schubert S, Weller U, Boogen C, Kostrzewa M.
Matrix-assisted laser desorption ionization-time of flight mass
spectrometry-based functional assay for rapid detection of
resistance against beta-lactam antibiotics. J Clin Microbiol.
2012;50:92737. doi:10.1128/jcm.05737-11.
39. Mitteregger D, Barousch W, Nehr M, Kundi M, Zeitlinger M,
Makristathis A, etal. Neutralization of antimicrobial substances
in new BacT/Alert FA and FN Plus blood culture bottles. J Clin
Microbiol. 2013;51:153440. doi:10.1128/jcm.00103-13.
Copyright of Infection is the property of Springer Science & Business Media B.V. and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.