Am J Respir Crit Care Med Vol 164. pp 396–402, 2001Internet address: www.atsjournals.org

To assess the diagnostic value of procalcitonin (PCT), interleukin(IL)-6, IL-8, and standard measurements in identifying critically illpatients with sepsis, we performed prospective measurements in78 consecutive patients admitted with acute systemic inflamma-tory response syndrome (SIRS) and suspected infection. We esti-mated the relevance of the different parameters by using multi-variable regression modeling, likelihood-ratio tests, and area underthe receiver operating characteristic curves (AUC). The final diag-nosis was SIRS in 18 patients, sepsis in 14, severe sepsis in 21, andseptic shock in 25. PCT yielded the highest discriminative value,with an AUC of 0.92 (CI, 0.85 to 1.0), followed by IL-6 (0.75; CI,0.63 to 0.87), and IL-8 (0.71; CI, 0.59 to 0.83; p

�

0.001). At a cut-off of 1.1 ng

�

ml, PCT yielded a sensitivity of 97% and a specificityof 78% to differentiate patients with SIRS from those with sepsis-related conditions. Median PCT concentrations on admission (ng

�

ml, range) were 0.6 (0 to 5.3) for SIRS; 3.5 (0.4 to 6.7) for sepsis;6.2 (2.2 to 85) for severe sepsis; and 21.3 (1.2 to 654) for septicshock (p

�

0.001). The addition of PCT to a model based solely onstandard indicators improved the predictive power of detectingsepsis (likelihood ratio test; p

�

0.001) and increased the AUCvalue for the routine value-based model from 0.77 (CI, 0.64 to0.89) to 0.94 (CI, 0.89 to 0.99; p

�

0.002). In contrast, no additiveeffect was seen for IL-6 (p

�

0.56) or IL-8 (p

�

0.14). Elevated PCTconcentrations appear to be a promising indicator of sepsis innewly admitted, critically ill patients capable of complementingclinical signs and routine laboratory parameters suggestive of se-vere infection.

Keywords:

Critical care; biological markers,

blood; calcitonin,

blood;

protein precursors,

blood; interleukin-6,

blood; interleukin-8,

blood;sepsis,

blood,

diagnosis; sepsis syndrome,

blood,

diagnosis

Sepsis can be difficult to distinguish from other noninfectiousconditions in critically ill patients admitted with clinical signsof acute inflammation (1, 2). This issue is of paramount impor-tance given that therapies and outcomes greatly differ between

patients with and those without sepsis. Thus, there is an unmetneed for clinical or laboratory tools distinguishing betweenthe systemic inflammatory response syndrome (SIRS) and thevarious forms of sepsis. To date, no single clinical or biologicalindicator of sepsis has gained unanimous acceptance (3), al-though several attempts have been made to correlate cytokinelevels with sepsis and patient prognosis (1, 4). Among the po-tentially useful sepsis markers, interleukin (IL)-6, IL-8, andprocalcitonin (PCT) have been proposed to be the most prom-ising candidates (1). The latter in particular has been postu-lated to be superior to clinical variables or commonly usedlaboratory tests, such as C-reactive protein or leukocyte count,and to even correlate with the severity of microbial invasion(5–9). However, several investigators have questioned the di-agnostic and prognostic accuracy of routine PCT measure-ments, retrieving inconsistent and variable results dependingon the severity of illness and infection in the studied patientpopulation (10–14).

In view of these conflicting findings and the paramount im-portance of the timely diagnosis of sepsis at time of admissionto the intensive care unit (ICU), the present cohort study pro-spectively investigated the diagnostic value of PCT, IL-6, andIL-8 in a group of severely ill patients admitted with signs ofacute, severe inflammation. Specifically, we assessed whetherPCT, IL-6, IL-8, or standard parameters at time of admissionare helpful in distinguishing sepsis from severe systemic non-infectious inflammation in newly admitted, critically ill pa-tients with suspected infection.

METHODS

Study Setting and Population

Over a 7-mo period, all consecutive patients admitted with a sus-pected diagnosis of infection and hospitalized in the medical and sur-gical ICUs of the University of Geneva Hospitals were eligible, in-cluding neutropenic and immunosuppressed patients. On average,1,700 adult patients are admitted each year to the 18-bed medical ICUfor a mean length of stay of 4 d. The surgical ICU is a 20-bed unit with1,600 admissions and a mean length of stay of 4 d.

At time of enrollment, all included patients had to be newly admit-ted to one of the ICUs, had to have a clinically suspected infection,and had to fulfill at least two criteria of SIRS (15). Clinically sus-pected infection was defined as an explicit statement by the attendingphysician indicating the suspicion of an ongoing infection, combinedwith the initiation of a diagnostic work-up to rule out infection andthe prescription of antimicrobial therapy (16). Patients were not en-rolled in case of early discharge or death, withholding of life-support,or complete absence of antimicrobial treatment.

Every effort was made to identify patients with suspected sepsis asearly as possible after ICU admission. This was accomplished by dedi-cated research fellows, who visited the ICUs twice daily, and by thecooperation of the attending physicians, who indicated to the researchteam each patient admitted with suspected sepsis. Patients originatedeither from the emergency room, the general wards, or from the oper-ating room.

(

Received in original form September 19, 2000 and in revised form March 30, 2001

)

Supported by grants from the Swiss National Foundation for Scientific Research(Cooperation in Science and Research with CEEC

�

NIS, and Grant No1 32-50764to J.P.)

Dr. Pugin is the recipient of a fellowship from the Prof. Dr. Max Cloetta Foundation.

Dr. Harbarth is the recipient of a Max Kade Fellowship Award.

Dr. Holeckova is the recipient of a grant from the University of Geneva.

Presented in part at the 38th Interscience Conference on Antimicrobial Agentsand Chemotherapy, San Diego, California, September 1998.

The hospital ethics committee approved the study protocol.

Correspondence and requests for reprints should be addressed to Jérôme Pugin,M.D., Division of Medical Intensive Care, The University of Geneva Hospitals,1211 Geneva 14, Switzerland. E-mail: jerome.pugin@medecine.unige.ch

This article has an online data supplement, which is accessible from this issue’stable of contents online at www.atsjournals.org

Diagnostic Value of Procalcitonin, Interleukin-6,and Interleukin-8 in Critically Ill Patients Admittedwith Suspected Sepsis

STEPHAN HARBARTH, KATARINA HOLECKOVA, CÉLINE FROIDEVAUX, DIDIER PITTET, BARA RICOU,GEORGES E. GRAU, LASZLO VADAS, JÉRÔME PUGIN, and the Geneva Sepsis Network

Infection Control Program, Division of Infectious Diseases, Division of Medical Intensive Care, Division of Surgical Intensive Care, Department of Pathology, and Laboratoire Central de Chimie Clinique, The University of Geneva Hospitals, Geneva, Switzerland

Harbarth, Holeckova, Froidevaux,

et al.

: Procalcitonin, IL-6, and IL-8 for Sepsis Diagnosis 397

Data Collection

Variables recorded at baseline and several times daily during follow-up included patient demographics, principal diagnosis, vital signs, res-piratory parameters, routine blood tests and microbiologic culture re-sults. Survival or death in the ICU was assessed during a follow-up ofas long as 28 d. Death was defined as sepsis-related or sepsis-unre-lated (e.g., cerebral death or cardiogenic shock). The severity of thepatients’ condition was measured according to the simplified acutephysiology scoring (SAPS) II system (17). Microbiologic tests and an-timicrobial therapy were prescribed by physicians on service accord-ing to the usual practice in the participating ICUs, without interfer-ence by the research team. The hospital ethics committee on humanresearch approved the study protocol.

Included patients were classified as having SIRS, sepsis, severesepsis, or septic shock at the time of admission, according to estab-lished consensus definitions (15, 16). Importantly, this case ascertain-ment was done retrospectively by two independent investigators with-out the knowledge of plasma PCT, IL-6, and IL-8 values, on the basisof the review of the complete patient charts, results of microbiologiccultures, chest radiographs, and, when available, postmortem exami-nation reports. Concordance among the two independent investiga-tors was excellent and the interrater reliability was high (

�

�

0.9). Dis-agreement was settled by consensus involving a third party.

Measurements of Procalcitonin, IL-6, and IL-8 Plasma Levels

Within 12 h after admission and study entry, whole heparinized bloodwas drawn via an arterial line for cytokine measurements and dailythereafter at 6:00

A

.

M

. during the entire ICU stay until ICU dischargeor death. Blood was put on ice and plasma was collected by centrifu-gation at 4

�

C, aliquoted, and stored at

�

70

�

C until the day of assay.PCT, IL-6, and IL-8 plasma levels were determined batchwise usingcommercial assays. Plasma PCT concentrations were measured in du-plicates using an immunoassay with a sandwich technique and achemiluminescent detection system, according to the manufacturer’sprotocol (LumiTest; Brahms Diagnostica, Berlin, Germany). As re-cently pointed out by Müller and colleagues (9), the LumiTest doesnot measure exclusively the procalcitonin molecule, but also severalother calcitonin precursors that are “captured” by the sandwich ELISA.The term “procalcitonin” as used here may therefore lack scientificprecision, but it was favored over “calcitonin precursors” because it iswidely used by critical care physicians and investigators.

PCT levels were found to be

�

0.2 ng

�

ml in plasma from 15 healthydonors. The coefficient of variation (CV) of the test was 9% at a PCTplasma concentration of 1.2 ng

�

ml, and 4.4% at 56 ng

�

ml.IL-6 and IL-8 were measured using semiautomated, chemilumi-

nescent immunoassays (ImmuliteOne; DPC-Biermann, Bad Nauheim,Germany). The median value (range) of plasma IL-6 and IL-8 levelsin 15 healthy blood donors was 2.6 (

�

1 to 11.3) pg

�

ml and 5 (4 to 27)pg

�

ml, respectively. The CVs were 4.3 and 8.8%, respectively. PCT,IL-6, and IL-8 plasma measurements were performed by trained labo-ratory technicians, blinded to the patients’ clinical course. In a ran-domly selected subgroup of 35 patients, we also quantified plasma lev-els of TNF-

�

(Medgenix, Belgium), soluble TNF receptors (sTNFR) Iand II, IL-10, and IL-1ra, using ELISA tests (R&D Systems, Abing-don, UK).

Statistical Analysis

Values were expressed as the mean

SD or as the median and inter-quartile range in case of a skewed distribution. Comparison of groupdifferences for continuous variables was done by one-way ANOVAor nonparametric Kruskal-Wallis test, when appropriate. The signifi-cance of difference in proportions was tested with use of the

2

-statis-tic. We also used univariate-correlation analyses to establish the rela-tion between parameters.

Sensitivity, specificity, and positive and negative predictive valuesof each cytokine parameter were calculated according to standardmethods (18). The diagnostic accuracy of the cytokine plasma levelsdetermined at study entry was expressed as the area under the receiveroperating characteristic curve (AUC), which was derived from logisticregression analysis (19). These values were calculated for the cutoffthat represented the best discrimination as derived from the AUCs.For the purpose of the logistic regression analysis, which requires bi-

nary outcome events, case subjects classified as confirmed sepsis, se-vere sepsis, or septic shock (

�

sepsis syndrome) were compared withcontrol patients with SIRS and initial suspicion of infection.

We analyzed associations between different levels of each individ-ual clinical and biological parameter and the risk of sepsis syndromeat admission, after dividing the values for each marker into quartiles.This was done because few continuous variables satisfied the linearityassumption. To estimate the potential clinical relevance of the cyto-kine measurements, we used likelihood-ratio tests to determinewhether logistic regression models that included measurements of cy-tokines and routine clinical and laboratory variables provided a signif-icantly better fit than did logistic regression models limited to stan-dard laboratory and clinical measurements alone (20). In order todetermine the prognostic power of the candidate parameters regard-ing patients’ survival, we used time-dependent Cox proportional haz-ards modeling (21).

Statistical software programs (STATA, version 6.0, Stata Inc.;Statview, version 4.1, Abacus Concepts, Berkeley, CA) were used fordata processing and analyses. All p values were two-sided, and statisti-cal significance was set at an

�

-value of 0.05.

Figure 1. Plasma levels of PCT, IL-6, and IL-8 in patients with SIRS (n �18), sepsis (n � 14), severe sepsis (n � 21), and septic shock (n = 25).Data are presented as box plots with median lines, 25- and 75-percen-tile boxes, and 10- and 90-percentile error bars, using a log scale for theY-axis. The circles represent the outliers.

398

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 164 2001

RESULTS

Characteristics of the Study Population

We prospectively enrolled 78 adult ICU patients meeting thecriteria for SIRS (n

�

18), sepsis (n

�

14), severe sepsis (n

�

21), or septic shock (n

�

25). The main patient characteristicsare shown in Table E1 (

see

Online Data Supplement). As ex-pected, patients admitted with septic shock were more se-verely ill at baseline than were those without. The formerwere more likely to have a higher SAPS II score, heart rate,and lactate and to have a lower core temperature, mean arterialpressure, platelet count, arterial pH, and oxygenation status.Infections were microbiologically proven in 44 of 60 infectedpatients (73%) with 53% Gram-negative, 38% Gram-positivebacteria, and 9% mixed infections. Leading sources of infec-tion were the respiratory tract (n

�

38) and the intra-abdominalspace (n

�

10). Twenty-three infected patients had a docu-mented bloodstream infection.

Baseline Plasma Levels of PCT, IL-6, and IL-8

Baseline plasma levels of the inflammation markers IL-6, IL-8,and PCT were higher among case subjects with severe sepsisor septic shock than among control subjects with SIRS only(Figure 1). PCT appeared to be most helpful in differentiatingpatients with sepsis from those with SIRS. Median PCT levelson admission (ng

�

ml, range) were 0.6 (0 to 5.3) for SIRS; 3.5(0.4 to 6.7) for sepsis, 6.2 (2.2 to 85) for severe sepsis; and 21.3(1.2 to 654) for septic shock (p

�

0.001). The accuracy of thecandidate parameters to distinguish patients with SIRS fromthose with septic conditions varied (Table 1). As shown in Fig-ure 2 (

upper panel

), PCT yielded the highest discriminativevalue with an AUC of 0.92 (95% confidence interval [CI], 0.85to 1.0) followed by IL-6 (AUC, 0.75; CI, 0.63 to 0.87) and IL-8(AUC, 0.71; CI, 0.59 to 0.83; p

�

0.001). At a cutoff point setat 1.1 ng

�

ml, PCT yielded a sensitivity of 97% and a specificityof 78% to differentiate patients with SIRS from those withsepsis, severe sepsis, or septic shock.

Serum C-reactive protein concentrations increased from119

89 mg

�

l for SIRS to 159

51 mg

�

L for sepsis, 254

181mg

�

l for severe sepsis, and 228

119 mg

�

l for septic shock(p

�

0.04 in Table E1) (

See

Online Data Supplement). C-reac-tive protein yielded an AUC of 0.76 (CI, 0.58 to 0.93), a valuesimilar to IL-6. At a cutoff value of 150 mg

�

l, the sensitivityand specificity were 68 and 73%, respectively.

Correlation of the Study Parameters with OtherSepsis Mediators

In a randomly selected subgroup of 35 patients, additionalproinflammatory and antiinflammatory cytokines (TNF-

�

,STNFR I and II, IL-10, and IL-1ra) were determined and cor-related with the main study parameters PCT, IL-6, and IL-8.

Several of those cytokines levels were highly correlated (Ta-ble E2.

See

Online Data Supplement). For example, the corre-lation coefficient for the relation between PCT and TNF-

�

was 0.86; between PCT and sTNFR I, 0.78; and between IL-6and TNF-

�

, 0.92 (all p

�

0.001). In contrast, correlation coeffi-cients between routinely available clinical and laboratory indi-cators of inflammation (e.g., leukocyte count, C-reactive pro-tein, temperature) and the additionally measured cytokineswere small; less than 40% of the variance in any of the cyto-kine markers was explained by any of the standard measures.

Discriminative Power of Clinical and BiologicalSepsis Indicators

In an additional attempt to assess the diagnostic value of thesepsis indicators, we stratified the continuous values for eachindicator into quartiles, and then analyzed associations be-tween different levels of each individual clinical and biologicalmarker and the risk of sepsis syndrome at admission (TableE3.

See

Online Data Supplement). No differences in clinicalvariables such as temperature or heart rate were observed be-tween septic patients and noninfected patients with SIRS, ex-cept for the mean arterial pressure (p

�

0.03). Conversely, sixbiological parameters predicted sepsis syndrome by bivariateanalysis (Table E3). In order of significance, these were PCT,IL-6, IL-8, monocyte count, arterial pH, and C-reactive pro-tein. PCT was the most powerful predictor of sepsis syndromein this bivariate analysis (p

�

0.001).

TABLE 1. DIAGNOSTIC PERFORMANCE OF DIFFERENTSEPSIS INDICATORS

Procalcitonin Interleukin-6 Interleukin-8

Cutoff value, ng/ml* 1.1 200 30Sensitivity, % 97 67 63Specificity, % 78 72 78Positive predictive value, % 94 89 90Negative predictive value, % 88 39 39Area under the receiver operating 0.92 0.75 0.71curve (95% confidence interval) (0.85–1.0) (0.63–0.87) (0.59–0.83)

* Sensitivity, specificity, and predictive values were calculated for the cutoff, whichrepresented the best discrimination as derived from the receiver operating characteris-tic curves.

Figure 2. Receiver operating characteristics curves (ROC) of plasmaparameters (PCT; IL-6; and IL-8; upper panel), and of a clinical modelwith and without PCT (lower panel). Areas under the ROC were: upperpanel, PCT, 0.92; IL-6, 0.75; IL-8, 0.71; and lower panel, clinical modelwith PCT, 0.94, and clinical model without PCT, 0.77.

Harbarth, Holeckova, Froidevaux,

et al.

: Procalcitonin, IL-6, and IL-8 for Sepsis Diagnosis 399

Clinical Significance of the Diagnostic Indicators

To explore the diagnostic performance of the parameters froma clinical perspective, we evaluated the accuracy of the cyto-kine levels and clinical indicators of inflammation as predic-tors of sepsis in a series of analyses. First, we determined theindependent predictive value of the different clinical and bio-logical sepsis indicators by performing multivariable logisticregression modeling. In the global stepwise analysis includingall variables described in Table E3, only PCT (adjusted oddsratio [AOR] per 1-point increase, 2.3; CI, 1.4 to 3.7; p

�

0.001)was found to be an independent predictor of sepsis syndrome.In a similar model that was limited to the routinely availableparameters without PCT, IL-6, and IL-8, increasing levels ofC-reactive protein (AOR, 1.01; CI, 1.0 to 1.03; p

�

0.023) andan elevated proportion of leukocyte band forms (AOR, 1.06;CI, 1.0 to 1.14; p

�

0.067) tended to predict sepsis syndrome.Second, likelihood-ratio tests were used to compare the fit

of predictive models that were based on measurement of rou-tinely available sepsis markers in combination with the studyparameters to the fit of models based only on standard indica-tors (core temperature, heart rate, blood pressure, white bloodcell count). In these models, the addition of PCT significantlyimproved the power of detecting sepsis syndrome (p

�

0.001).

In contrast, no additive effect was seen for IL-6 (p

�

0.56) orIL-8 (p

�

0.14).Third, as a measure of clinical usefulness, we evaluated the

combined role of the cytokine levels and clinical markers aspredictors of sepsis using AUC analysis. As shown in Figure 2(

lower panel

), the inclusion of PCT increased the AUC valuefor the routine value-based model from 0.77 (CI, 0.64 to 0.89)to 0.94 (CI, 0.89 to 0.99; p

�

0.002). Again, no such effect wasobserved with IL-6 or IL-8.

Severity of Sepsis and Final Outcome

We further evaluated the time course of PCT, IL-6, and IL-8in relation to the prognosis and severity of sepsis. The evolu-tion of serum PCT concentrations during follow-up of patientswith severe sepsis and septic shock are shown in Figure 3. Aslow decrease or no decrease in PCT levels 48 h after admis-sion was consistently associated with a poor outcome (Figure3,

panel B

). All patients who died of sepsis or related compli-cations never had PCT levels below 1.1 ng

�

mL, except for onesingle time point in one patient. Conversely, PCT levels de-creased under the cutoff of 1.1 ng

�

ml within 8 d in all survivorsexcept for one (Figure 3,

panel A

). In contrast, IL-6 and IL-8did not demonstrate this uniform pattern in septic patients(Figure E1.

See

Online Data Supplement). Within the 8-d pe-riod after onset of severe sepsis or septic shock, six survivorsstill had elevated IL-6 or IL-8 levels. Interestingly, we ob-served that in four patients with prolonged ICU stay and com-

Figure 3. Daily variations of procalcitonin (PCT) levels during ICU hos-pitalization. (Panel A) Patients admitted with severe sepsis and septicshock who survived. (Panel B) Patients admitted with severe sepsis andseptic shock who died. (Panel C) Control patients admitted with SIRSand initial suspicion of infection.

Figure 4. PCT, IL-6, and IL-8baseline levels in patients withsepsis, severe sepsis, and sep-tic shock who survived (n �38) or who died (n � 18)of sepsis-related death. Fourpatients who died from causesunrelated to sepsis wereomitted from the analysis.Data are presented as boxplots with median lines, 25-and 75-percentile boxes, and10- and 90-percentile errorbars, using a log scale forthe Y-axis. The circles rep-resent the outliers.

400

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 164 2001

plicated courses of sepsis, peaks in plasma PCT levels coin-cided with clinical episodes of septic shock (Figure 3,

panel D

)(Figure E2.

See

Online Data Supplement).Values of plasma PCT, IL-6, and IL-8 in infected patients

at the time of admission relative to sepsis-related death areshown in Figure 4. The most discriminative parameter to pre-dict sepsis-related death in infected patients at the time of ad-mission was an IL-6 value above 1,000 ng

�

ml.Finally, we assessed all follow-up values of the candidate

parameters as determinants of survival in the entire cohort ofinfected and noninfected patients. During 832 d at risk, themean plasma levels of IL-6 (mean difference, 4,689.1 ng

�

ml;p

�

0.0003), IL-8 (mean difference, 1021.9 ng

�

ml; p

�

0.0001)and PCT (mean difference, 25.7 ng�ml; p � 0.004) differed be-tween those patients who survived and those who died. Wethen included each cytokine measurement in a time-dependentCox proportional hazards model. In this longitudinal analysisthat also included noninfected patients, only increasing levelsof IL-8 predicted death (AOR per 1,000 ng increment, 1.05; CI,1.00 to 1.10; p � 0.046), whereas IL-6 (p � 0.82) and PCT (p �0.36) did not indicate an increased risk of death.

DISCUSSION

The principal findings of this study performed in newly admitted,critically ill patients with suspected sepsis are: (1) the confirma-tion of the previously described good diagnostic accuracy of PCT;(2) the modest discriminative value of IL-6, IL-8, and other indi-cators of inflammation; (3) the additive effect of PCT to improvethe predictive power of routinely available sepsis parameters; (4)the excellent correlation of PCT with other biologic markers ofsepsis such as TNF-�; and (5) the good concordance of PCTplasma levels with the clinical evolution of septic patients.

Critical care physicians have at their disposal a variety ofdata to serve as a guide in discriminating infectious from non-infectious conditions in newly admitted patients. In a numberof cases, the diagnosis of sepsis becomes clear after complet-ing the medical history and physical examination of a newlyadmitted patient (22). In other circumstances of non-infec-tious insults causing SIRS (e.g., trauma, burns, hemorrhages,hypothermia, pancreatitis, and surgery), or in comatose pa-tients, the diagnosis of sepsis remains difficult. In these cases,several indicators have been proposed as new diagnostic teststo assess the competing probabilities of various candidate ill-nesses in the differential diagnosis of sepsis (23, 24). The presentresults confirm and extend earlier findings demonstrating thatPCT is among the most promising sepsis markers in criticallyill patients, capable of complementing clinical signs and rou-tine laboratory parameters suggestive of severe infection attime of ICU admission. Recently, Müller and colleagues (9)investigated 101 patients admitted to a medical ICU and sug-gested that PCT is a more reliable marker of sepsis than isC-reactive protein, IL-6, and lactate levels. Another groupfrom Germany (25) recently reported average PCT concentra-tions of 0.4 3.4 ng�ml for SIRS, 0.5 2.9 ng�ml for sepsis,6.9 3.9 ng�ml for severe sepsis and 12.9 4.4 ng�ml for septicshock. On the basis of their findings, the investigators con-cluded that PCT measurement appears to be useful to discrim-inate between sepsis and severe sepsis, in contrast to C-reac-tive protein, blood cell counts, or body temperature. Lastly,Viallon and colleagues (26) investigated 61 patients with cir-rhotic ascites or spontaneous bacterial peritonitis and foundPCT values similar to those in the present study: median PCTserum levels and interquartile ranges were 3.2 (2.3 to 10.1),13.8 (8.7 to 18.4), and 26.5 (24.0 to 52.0) ng�ml in patients withsepsis, severe sepsis, and septic shock, respectively.

Our study has several important implications for clinicians.First, as a putative new test to diagnose sepsis on ICU admis-sion, PCT offers a high level of certainty than other currentlyavailable candidates do. In the current study, a cutoff point of1.1 ng�ml detected virtually all septic patients. This accuracy,although not perfect, may guide physicians in their clinical de-cision-making and their stepwise approach to the complexmanagement of critically ill patients requiring several interven-tions in a short period of time. Second, in the recovery phase ofsevere sepsis or septic shock, a greatly decreased PCT level in-dicates a favorable clinical evolution and may prompt physi-cians to withdraw antimicrobial treatment. Third, measure-ment of PCT is of practical value since it increases 2 to 4 h afterbacterial or endotoxin challenge (24). The test can be per-formed within 3 to 4 h and may give valuable information longbefore blood culture results are back. Thus, in contrast to manyother cytokine markers with purely theoretical utility (27),measuring PCT with a commercially available assay is realistic,inexpensive and may even be cost-effective by better directingdiagnostic and therapeutic interventions in critically ill patientswith SIRS of undetermined origin. This, of course, requiresfurther prospective testing. Finally, PCT is not only helpful inthe diagnosis of sepsis, but also indicative of its severity. Aspreviously suggested (9, 28), we demonstrated that constantlyelevated PCT levels in infected patients were associated withpoor prognosis. In contrast, the rapid decrease of PCT levelsduring the first week after ICU admission had a good prognos-tic value. The relation between these findings and the patho-physiology of sepsis remains to be elucidated. It is, however,tempting to hypothesize that persistently elevated PCT valuesduring the course of sepsis may identify subgroups of patientswho have difficulties in clearing their systemic infection.

One interesting finding was that PCT levels in infected pa-tients correlated well with circulating cytokines such as TNF-�,thus representing a suitable surrogate marker for key cytokinesduring septic shock. Taking into account that poor immuno-logic characterization was one of the main reasons for failure intrials with immunomodulating agents, the rapid measurementof PCT may improve the results of future intervention trials todetect and treat patients with severe sepsis and septic shock (8,29). Thus, bedside documentation of elevated PCT levels maybecome an additional entry criterion for sepsis trials (8).

Our study has several strengths. First, our study populationwas a large, diverse group of critically ill adult patients admit-ted to both medical and surgical ICUs in different phases ofinfectious and noninfectious conditions. This allowed us tomake group comparisons not previously reported by otherstudies, in order to increase the generalizability of the studyfindings. In contrast, most of the previous studies that at-tempted to draw conclusions about PCT were based on smallpatient samples (30) or selected patient groups such as febrilechildren, postsurgical patients, or patients with specific dis-eases only (31–34). Second, following high methodologic stan-dards (35), we used prospective surveillance and consecutivepatient enrollment, intensive clinical and laboratory data col-lection and follow-up, and powerful statistical methods to ac-curately describe and analyze the diagnostic performance ofthe studied parameters. Third, outcomes were determined byblinded investigators without knowledge of the plasma cyto-kine levels, and vice versa. Fourth, we classified patients withSIRS and suspected infection according to predetermined,well-established outcome categories, after incorporation of allavailable clinical, laboratory, and histologic evidence. Fifth, incontrast to other studies (9, 30, 31), we did not include clini-cally unrealistic control patients without suspected infection,but only patients with a high pretest probability of sepsis, cov-

Harbarth, Holeckova, Froidevaux, et al.: Procalcitonin, IL-6, and IL-8 for Sepsis Diagnosis 401

ering the spectrum of patients that is likely to be encounteredin the future use of the tests (35). Finally, our study was de-signed as a real-life study, to closely resemble clinical practice.Therefore, we evaluated the combined role of the cytokinesand clinical indicators of inflammation as predictors of sepsisin a series of analyses, in which we explored the diagnostic ac-curacy of these different parameters from a clinical perspec-tive. In summary, in contrast to many previous studies, ourstudy fulfills all important criteria for the optimal design ofstudies assessing the accuracy of diagnostic tests (35).

IL-6 and IL-8 are potent inflammatory mediators and theirplasma concentrations have been tested as prognostic factors inseveral studies (27, 36, 37). The results of the present study sug-gest that measurement of these cytokines is not optimal for dis-criminating patients with infectious from those with noninfec-tious conditions. In particular, we confirm the observation byother investigators that IL-8 is not a good indicator of systemicinfection, although elevated IL-8 levels are indicative of multi-ple organ failure and an increased likelihood of death (38). Ourfindings are also consistent with the results reported by DeWerra and colleagues (39), who observed increased IL-6 con-centrations in patients with cardiogenic shock. Thus, persis-tently high IL-6 and IL-8 levels are very sensitive markers of in-flammation and are likely to be of prognostic value in criticallyill patients with multiorgan failure (27, 40). However, their diag-nostic accuracy is limited in critically ill patients by the unspe-cific elevation caused by the accompanying inflammation inde-pendent of the infection. In contrast to IL-6 and IL-8, PCT doesrespond rather to an infection than to the inflammation.

Several limitations of this study merit consideration. Usingclinical criteria and microbiologic evidence, it may have beendifficult to ascertain the exact etiology of SIRS in all patients.This may have introduced some misclassification bias, butsince investigators blinded to the cytokine results performedthe case ascertainment, we do not believe that this lack of agold standard compromised our study conclusions. Our dataare also limited because we did not include enough patientswith underlying autoimmune disease or other diseases thatmay increase PCT levels without infection. Major surgery suchas kidney transplantation or cardiopulmonary bypass surgery,in particular, may distort PCT measurement (32, 41). More-over, antibiotic therapy may have an important impact onPCT values. Because antimicrobial treatment was a criterionfor study inclusion, all patients in our study already received atleast one dose of an antimicrobial agent at the time of firstblood collection. Thus, similar studies should be conducted inorder to elucidate the exact relation between adequate antibi-otic treatment and the time course of daily PCT serum levels.Finally, our results may not apply to patients with localized ormild infections not requiring ICU treatment. In those patients,PCT levels may fall below the proposed cutoff and other pa-rameters may be more useful to diagnose or exclude infection.

In conclusion, in this prospective, blinded evaluation ofthree different sepsis indicators in a consecutive series of pa-tients, PCT proved to be the best indicator of infection in newlyadmitted critically ill patients. Our results indicate that clinicalvariables are of moderate diagnostic value for the diagnosis ofsepsis on ICU admission. Thus, our data raise the possibilitythat the addition of PCT to the standard work-up of critically illpatients with suspected sepsis could increase diagnostic cer-tainty and generate an improved patient management.

References1. Carlet J. Rapid diagnostic methods in the detection of sepsis. Infect Dis

Clin North Am 1999;13:483–494.

2. Ueda S, Nishio K, Minamino N, Kubo A, Akai Y, Kangawa K, MatsuoH, Fujimura Y, Yoshioka A, Masui K, et al. Increased plasma levels ofadrenomedullin in patients with systemic inflammatory response syn-drome. Am J Respir Crit Care Med 1999; 160:132–136.

3. Boucher BA, Hanes SD. Searching for simple outcome markers in sep-sis: an effort in futility? Crit Care Med 1999;27:1390–1391.

4. Gogos CA, Drosou E, Bassaris HP, Skoutelis A. Pro- versus anti-inflam-matory cytokine profile in patients with severe sepsis: a marker forprognosis and future therapeutic options. J Infect Dis 2000;181:176–180.

5. Gendrel D, Raymond J, Coste J, Moulin F, Lorrot M, Guerin S, RavillyS, Lefevre H, Royer C, Lacombe C, et al. Comparison of procalcitoninwith C-reactive protein, interleukin 6 and interferon-alpha for differ-entiation of bacterial vs. viral infections. Pediatr Infect Dis J 1999;18:875–881.

6. Hatherill M, Tibby SM, Sykes K, Turner C, Murdoch IA. Diagnosticmarkers of infection: comparison of procalcitonin with C reactive pro-tein and leucocyte count. Arch Dis Child 1999;81:417–421.

7. Bossink AW, Groeneveld AB, Thijs LG. Prediction of microbial infec-tion and mortality in medical patients with fever: plasma procalcito-nin, neutrophilic elastase-alpha1-antitrypsin, and lactoferrin com-pared with clinical variables. Clin Infect Dis 1999;29:398–407.

8. Oberhoffer M, Karzai W, Meier-Hellmann A, Bogel D, Fassbinder J,Reinhart K. Sensitivity and specificity of various markers of inflam-mation for the prediction of tumor necrosis factor-alpha and interleu-kin-6 in patients with sepsis. Crit Care Med 1999;27:1814–1818.

9. Müller B, Becker KL, Schachinger H, Rickenbacher PR, Huber PR,Zimmerli W, Ritz R. Calcitonin precursors are reliable markers ofsepsis in a medical intensive care unit. Crit Care Med 2000;28:977–983.

10. Mimoz O, Benoist JF, Edouard AR, Assicot M, Bohuon C, Samii K.Procalcitonin and C-reactive protein during the early posttraumaticsystemic inflammatory response syndrome. Intensive Care Med 1998;24:185–188.

11. Lapillonne A, Basson E, Monneret G, Bienvenu J, Salle BL. Lack ofspecificity of procalcitonin for sepsis diagnosis in premature infants.Lancet 1998;351:1211–1212.

12. Ugarte H, Silva E, Mercan D, De Mendonca A, Vincent JL. Procalcito-nin used as a marker of infection in the intensive care unit. Crit CareMed 1999;27:498–504.

13. Ruokonen E, Nousiainen T, Pulkki K, Takala J. Procalcitonin concen-trations in patients with neutropenic fever. Eur J Clin Microbiol InfectDis 1999;18:283–285.

14. Vincent JL. Procalcitonin: THE marker of sepsis? Crit Care Med 2000;28:1226–1228.

15. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA,Schein RM, Sibbald W. Definitions for sepsis and organ failure andguidelines for the use of innovative therapies in sepsis. Chest 1992;101:1644–1655.

16. Rangel-Frausto SM, Pittet D, Costigan M, Hwang TS, Davis CS, WenzelRP. The natural history of the systemic inflammatory response syn-drome (SIRS): a prospective study. JAMA 1995;273:117–123.

17. Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiol-ogy Score (SAPS II) based on a European-North American multi-center study. JAMA 1993;270:2957–2963.

18. Farr BM, Shapiro DE. Diagnostic tests: distinguishing good tests frombad and even ugly ones. Infect Control Hosp Epidemiol 2000;21:278–284.

19. Hanley JA, McNeil BJ. The meaning and use of the area under a re-ceiver operating characteristic (ROC) curve. Radiology 1982;143:29–36.

20. Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive protein andother markers of inflammation in the prediction of cardiovascular dis-ease in women. N Engl J Med 2000;342:836–843.

21. Christensen E. Multivariate survival analysis using Cox’s regressionmodel. Hepatology 1987;7:1346–1358.

22. Bates DW, Cook EF, Goldman L, Lee TH. Predicting bacteremia inhospitalized patients: a prospectively validated model. Ann InternMed 1990;113:495–500.

23. Assicot M, Gendrel D, Carsin H, Raymond J, Guilbaud J, Bohuon C.High serum procalcitonin concentrations in patients with sepsis andinfection. Lancet 1993;341:515–518.

24. Karzai W, Oberhoffer M, Meier-Hellmann A, Reinhart K. Procalcito-nin: a new indicator of the systemic response to severe infections. In-fection 1997;25:329–334.

25. Brunkhorst FM, Wegschneider K, Forycki ZF, Brunkhorst R. Procalci-tonin for early diagnosis and differentiation of SIRS, sepsis, severesepsis, and septic shock. Intensive Care Med 2000;26:S148–S152.

26. Viallon A, Zeni F, Pouzet V, Lambert C, Quenet S, Aubert G, Guyo-march S, Tardy B, Bertrand JC. Serum and ascitic procalcitonin levelsin cirrhotic patients with spontaneous bacterial peritonitis: diagnostic

402 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 164 2001

value and relationship to pro-inflammatory cytokines. Intensive CareMed 2000;26:1082–1088.

27. Casey LC, Balk RA, Bone RC. Plasma cytokine and endotoxin levelscorrelate with survival in patients with the sepsis syndrome. Ann In-tern Med 1993;119:771–778.

28. Schroder J, Staubach KH, Zabel P, Stuber F, Kremer B. Procalcitonin asa marker of severity in septic shock. Langenbecks Arch Surg 1999;384:33–38.

29. Abraham E, Matthay MA, Dinarello CA, Vincent JL, Cohen J, OpalSM, Glauser M, Parsons P, Fisher Jr. CJ, Repine JE. Consensus con-ference definitions for sepsis, septic shock, acute lung injury, andacute respiratory distress syndrome: time for a reevaluation. Crit CareMed 2000;28:232–235.

30. Selberg O, Hecker H, Martin M, Klos A, Bautsch W, Kohl J. Discrimi-nation of sepsis and systemic inflammatory response syndrome by de-termination of circulating plasma concentrations of procalcitonin,protein complement 3a, and interleukin-6. Crit Care Med 2000;28:2793–2798.

31. Chiesa C, Panero A, Rossi N, Stegagno M, De Giusti M, Osborn JF,Pacifico L. Reliability of procalcitonin concentrations for the diagno-sis of sepsis in critically ill neonates. Clin Infect Dis 1998;26:664–672.

32. Meisner M, Tschaikowsky K, Hutzler A, Schick C, Schuttler J. Postoper-ative plasma concentrations of procalcitonin after different types ofsurgery. Intensive Care Med 1998;24:680–684.

33. Muller CA, Uhl W, Printzen G, Gloor B, Bischofberger H, TcholakovO, Buchler MW. Role of procalcitonin and granulocyte colony stimu-lating factor in the early prediction of infected necrosis in severe acutepancreatitis. Gut 2000;46:233–238.

34. Kettelhack C, Hohenberger P, Schulze G, Kilpert B, Schlag PM. Induc-tion of systemic serum procalcitonin and cardiocirculatory reactionsafter isolated limb perfusion with recombinant human tumor necrosisfactor-alpha and melphalan. Crit Care Med 2000;28:1040–1046.

35. Lijmer JG, Mol BW, Heisterkamp S, Bonsel GJ, Prins MH, van der

Meulen JH, Bossuyt PM. Empirical evidence of design-related bias instudies of diagnostic tests. JAMA 1999;282:1061–1066.

36. Patel RT, Deen KI, Youngs D, Warwick J, Keighley MR. Interleukin 6is a prognostic indicator of outcome in severe intra-abdominal sepsis.Br J Surg 1994;81:1306–1308.

37. Gardlund B, Sjolin J, Nilsson A, Roll M, Wickerts CJ, Wretlind B.Plasma levels of cytokines in primary septic shock in humans: correla-tion with disease severity. J Infect Dis 1995;172:296–301.

38. Rau B, Steinbach G, Gansauge F, Mayer JM, Grunert A, Beger HG.The potential role of procalcitonin and interleukin 8 in the predictionof infected necrosis in acute pancreatitis. Gut 1997;41:832–840.

39. De Werra I, Jaccard C, Corradin SB, Chiolero R, Yersin B, Gallati H,Assicot M, Bohuon C, Baumgartner JD, Glauser MP, et al. Cytokines,nitrite�nitrate, soluble tumor necrosis factor receptors, and procalcito-nin concentrations: comparisons in patients with septic shock, cardio-genic shock, and bacterial pneumonia. Crit Care Med 1997;25:607–613.

40. Damas P, Ledoux D, Nys M, Vrindts Y, De Groote D, Franchimont P,Lamy M. Cytokine serum level during severe sepsis in human IL-6 asa marker of severity. Ann Surg 1992;215:356–362.

41. Eberhard OK, Langefeld I, Kuse ER, Brunkhorst FM, Kliem V, SchlittHJ, Pichlmayr R, Koch KM, Brunkhorst R. Procalcitonin in the earlyphase after renal transplantation: will it add to diagnostic accuracy?Clin Transplant 1998;12:206–211.

Appendix

Members of the Geneva Sepsis Network involved in thepresent study were: D. Pittet (chair), J.-C. Chevrolet, J.-M.Dayer, P. Eggimann, J. Garbino, G. E. Grau, S. Harbarth, D.Hochstrasser, A. Kocher, D. Lew, C. Pastor, J. Pugin, B. Ricou,J. A. Romand, N. Sotirov, P. M. Suter.