Received: 21 August 2018 | Accepted: 19 September 2018

DOI: 10.1002/jcb.27870

RESEARCH ARTICLE

The diagnostic accuracy of procalcitonin and C‐reactive

protein for sepsis: A systematic review and meta‐analysis

Meichun Tan | Yunxia Lu | Hao Jiang | Liandong Zhang

Department of Emergency (EICU),

Shanghai Baoshan Traditional Chinese Abstract
Medicine—Integrated Hospital, Shanghai, Background: The objective of this study was to systematically evaluate the
China
clinical value of procalcitonin and C‐reactive protein in the diagnosis of adult
Correspondence patients with sepsis.
Liandong Zhang, Department of Method: PubMed, Cochrane, Embase, Wanfang, China National Knowledge
Emergency (EICU), Shanghai Baoshan
Infrastructure, and VIP database were searched by the index words to identify
Traditional Chinese Medicine—
Integrated Hospital, Shanghai 201999, the qualified prospective studies, and relevant literature sources were also
China. searched. The most recent research was done in the April 2017. The only
Email: tanmeichun_2@126.com
languages included were English or Chinese. In the experiment group, patients
were diagnosed with sepsis, severe sepsis, or septic shock; in the control group,
the patients were of noninfectious origin or a systemic inflammatory response
syndrome. The diagnostic accuracy was analyzed by heterogeneity, diagnostic
odds ratio (DOR), sensitivity, specificity, positive likelihood ratio, negative
likelihood ratio, and the summary receiver‐operating characteristic curve.
Results: At least nine studies were involved in the meta‐analysis with 495
patients in the sepsis group and 873 patients in the nonsepsis group. In terms of
the diagnostic accuracy of C‐reactive protein (CRP) for sepsis, the overall area
under the summary receiver operator characteristic (SROC) curve was 0.73
(95% confidence interval [CI], 0.69‐0.77), with a sensitivity and specificity of
0.80 (95% CI, 0.63‐0.90) and 0.61 (95% CI, 0.50‐0.72) respectively, and the DOR
was 6.89 (95% CI, 3.86‐12.31). In terms of the diagnostic accuracy of
procalcitonin (PCT) for sepsis, the overall area under the SROC curve was
0.85 (95% CI, 0.82‐0.88), with a sensitivity and specificity of 0.80 (95% CI, 0.69‐
0.87) and 0.77 (95% CI, 0.60‐0.88) respectively, and the DOR was 12.50 (95% CI,
3.65‐42.80).
Conclusion: In this meta‐analysis, our results together indicate a moderate
degree of value of PCT and CRP for the diagnosis of sepsis in adult patients. The
diagnosis accuracy and specificity of PCT are higher than those of CRP.

KEYWORDS
C‐reactive protein (CRP), meta‐analysis, procalcitonin, sepsis

Abbreviations: CRP, C‐reactive protein; PCT, procalcitonin; SIRS, systemic inflammatory response syndrome; SROC, summary receiver operator
characteristic.
J Cell Biochem. 2018;1-8. wileyonlinelibrary.com/journal/jcb © 2018 Wiley Periodicals, Inc. | 1
2 | TAN ET AL.

1 | INTRODUCTION 3 | I N C L U S I O N AN D E XC L U S I O N
C R I T E RI A
Sepsis is a life‐threatening condition that arises when the
body’s response to an infection injures its own tissues and Studies were included on the basis of the following criteria:
organs.1 The consensus conference of American College of (1) clinical trial studies (prospective, retrospective, cross‐
Chest Physicians and Society of Critical Care Medicine sectional, and cohort study); (2) the research subjects were
confirms that sepsis is a Systemic Inflammatory Response adult patients and in the experiment group, the patients
Syndrome (SIRS) that is caused by infection.2 Despite were diagnosed with sepsis, severe sepsis, or septic shock;
advances in antibiotic therapy and modern life support, the in the control group, the patients were noninfectious origin
fatality rate of patients with sepsis has remained as high as or a SIRS; (3) diagnostic criteria: the gold diagnostic criteria
30% to 60% worldwide. Severe sepsis or septic shock are formulated by ACCP or SCCM; (4) only included English
common in the ICU, with a high mortality rate of 30% to and Chinese articles; (5) obtained the true positive value,
60%.3,4 Studies have shown that the high mortality rate is not false positive values, true negative value, false negative
only related to the severity of the illness, but also to whether values of procalcitonin and C‐reactive protein in the
the disease could be diagnosed early. Early identification of diagnoses of sepsis.
patients at a high risk of dying from sepsis may help initiate A study was excluded if it was: (1) a repeat of published
rapid and appropriate therapeutic interventions and may articles (the content or the result were same); (2) data had
decrease the morbidity and mortality caused by sepsis.5,6 obvious mistakes or were incomplete;(3) case report,
Blood cultures are the gold standard to diagnose infection, theoretical research, conference report, systematic review,
but only 30% blood cultures of sepsis patients are positive.7,8 meta‐analysis, expert comment, economic analysis; and
The early clinical manifestations of sepsis are nonspecific, (4) the outcomes were not what we needed.
including fever, tachycardia, leukocytosis, and so on. The All the studies were screened by two reviewers
symptoms overlap with SIRS.9 The other symptoms of sepsis, independently to determine whether they satisfied the
such as hypotension and thrombocytopenia, appear later. A inclusion and exclusion criteria, discrepancies were
delayed diagnosis could result in a serious condition, resolved by involved a third reviewer.
multiple system organ failure, extended hospitalization time,
and increased mortality. Nonspecific inflammation indexes,
such as procalcitonin (PCT) and C‐reactive protein (CRP), 4 | DATA EXTRACTION AND
have been widely used in the clinical setting to identify QU ALITY ASSESSMENT
infections. PCT is a sugar protein that contains 116 amino
acids. It presents advantages such as wide biological The data for analyses were extracted from all the included
characteristics, short induction time of bacteria stimulation, studies and consisted of two parts: basic information and
and a long half‐life. These advantages help to diagnose sepsis main outcomes. The first part included the author name,
and evaluate its seriousness.10,11 the year of publishing, the sample size, the percentage of
We performed this systematic review and meta‐ male, the main age, the cutoff value of procalcitonin, and
analysis with the aim to investigate the diagnostic C‐reactive protein, and the relative content of ROC.
accuracy of PCT and CRP of sepsis based on prospective The second part included the clinical outcomes: the true
studies or randomized control trails. positive values, false positive values, true negative values,
false negative values, sensitivity, specificity of procalcitonin
and C‐reactive protein. All these processes were done by
2 | METHODS
two reviewers independently, and disagreements between
reviewers were resolved by discussions until a consensus
2.1 | Search strategy
was reached. And weighted kappa coefficients were utilized
The Cochrane, Pubmed, Embase, China National Knowl- for quality assessment.
edge Infrastructure, WanFang, Weipu (VIP) ScienceChina,
Intute, Springer, Blackwell, Ingenta, Kluwer, OVID, Pro-
Quest, Wiley InterScience, IEEE, EBSCO, ESI, and other 5 | STATI S TIC AL ANALYSI S
databases were searched for all the eligible studies, including
randomized controlled trials and prospective studies. Other All statistical analyses were performed in the STATA 10.0
related articles and reference materials were also searched. (STATA 10.0 Software, Inc. La Jolla). Chi‐squared and I2
The most recent research was performed on April 2017. Two tests were used to test the heterogeneity of the clinical
investigators searched the literature independently, and a trial results, and the analysis model (fixed‐effect model or
third investigator was involved when a discrepancy occurred. random‐effect model) was decided. Chi‐squared test P
TAN ET AL. | 3

SIRS (2); severe systematic infection vs without systemic

infection (1); and sepsis vs infected without sepsis (1).
The cutoff values of PCT ranged from 0.76 to 6.03 ng/mL.
The cutoff values of CRP ranged from 12.00 to 90.00 mg/
L. The cutoff values of PCT or CRP varied between each
study. A total of 1368 patients were included in this meta‐
analysis. The other basic information including main age
and the ratio of the male/female number in each study is
also presented in Table 1.

7 | THE DIAGNOSTIC
A C C UR A C Y O F C R P F OR S E P S IS
FIGURE 1 Flow diagram of the literature search and selection
process
The studies included 495 patients in the sepsis group and
873 patients in the nonsepsis group who underwent CRP
values of ≤0.1 and I2 test values >50% were defined as testing in practice and were clearly diagnosed with sepsis.
acceptable heterogeneity and assessed by a random The Deek’s funnel plot of the included studies suggested
effects model. Chi‐squared test P values of >0.1 and I2 that there was no significant publication bias of CRP
tests values ≤50% were defined as homogeneous data and diagnostic outcomes (Figure 2, P = 0.32).
assessed by a fixed effects model. Random effects model No statistically significant difference was observed when
or fixed effects model was used to calculate the pooled exploring for the threshold effect (Spearman correlation
sensitivity, specificity, diagnostic odds ratio (DOR), coefficient = 0.483, P = 0.188). The pooled sensitivity and
positive likelihood ratio, and negative likelihood ratio. specificity of CRP were 0.80 (95% confidence interval [CI],
We constructed a summary receiver operator character- 0.63‐0.90) and 0.61(95%,CI, 0.50‐0.72) respectively in Figure
istic (SROC) curve by plotting the individual and 3. The DOR was 6.89(95% CI, 3.86‐12.31) (Figure 4). The
summary points of sensitivity and specificity to assess overall area under the SROC curve of CRP was 0.73 (95% CI,
the overall diagnostic accuracy. Publication bias was 0.69‐0.77) (Figure 5).
evaluated by Deek’s funnel plot.

8 | THE DIAGNOSTIC
6 | RESULTS
A C C UR A C Y O F P C T F O R SE PS IS
6.1 | Characteristics of the included
The studies included 495 patients in the sepsis group and
studies
873 patients in the nonsepsis group who underwent PCT
In total, 615 articles were searched by the indexes, and testing in practice and were clearly diagnosed with sepsis.
548 articles were excluded by screening the title and the The Deek’s funnel plot of the included studies suggested
abstract, leaving 67 articles for further evaluation. After there was no significant publication bias of PCT
obtaining and thorough reviewing the full manuscripts, it diagnostic outcomes (Figure 6, P = 0.21).
was concluded that 58 articles did not meet the inclusion No statistically significant difference was observed
criteria because there were no outcomes or data when exploring for the threshold effect (Spearman
deficiency (13), subjects were newborn (16), unqualified correlation coefficient = −0.335, P = 0.344). The pooled
grouping (21), and unqualified content (8). At last, nine sensitivity and specificity of PCT were 0.80 (95% CI, 0.69‐
studies12-20 were involved in the meta‐analysis with 495 0.87) and 0.77 (95% CI, 0.60‐0.88), respectively, as shown
patients in the sepsis group and 873 patients in the in Figure 7. The DOR was 12.50 (95% CI, 3.65‐42.80)
nonsepsis group. The detailed selection process is (Figure 8). The overall area under the SROC curve of PCT
presented in Figure 1. was 0.85 (95% CI, 0.82‐0.88) (Figure 9).
The main characteristics of the included studies are
summarized in Table 1. The study designs of the included
studies included retrospective study (1), prospective study 9 | DISCUSSION
(5), cross‐sectional study (2), and cohort study (1). The
read out counter (ROC) comparison content included: In this meta‐analysis, we aimed to investigate the
sepsis vs critical ill (1); sepsis vs nonsepsis (4); sepsis vs accuracy of the diagnosis of PCT and CRP for adult
4 | TAN ET AL.

62.0 56.0
43.6 48.2

40.8 40.0

54.0 51.0
42.0 42.0

69.0 61.0
38.5 44.0
C

‐
Age

59.2
T

‐
32
28
29

29

29
29
29
43
11
C
Male

20
20
22

23

26
28
29
37
20
T
No. of patients

22
37
15

26

26
11
532
193
11
C
33
27
34

26

26
29
187
107
26
FIGURE 2 Publication bias of CRP. CRP, C‐reactive protein
T

sepsis. Currently, blood culture is the gold standard in the

Cutoff values of

diagnosis of sepsis, but there are some restrictions such as

CRP, mg/L

poor sensitivity and longer inspection time. The defini-

tions of sepsis or septic shock have reached an interna-
Abbreviations: CRP, C‐reactive protein; PCT, procalcitonin; ROC, read out counter; SIRS, systemic inflammatory response syndrome.
50.70
61.75
90.00

21.30

16.50
23.00
12.00

65.00

tional consensus, but there is a great difference in the

clinical diagnosis for the doctor experience or the
‐

complexities of the disease. Therefore, we need labora-

Cutoff values of

tory examination (such as PCT, CRP, TNF, et al) to

PCT, ng/mL

complement the diagnosis of sepsis. PCT has a stable

existence in blood indexes; it is easy to test, and the
1.10
6.03
1.20

1.16

2.10
0.57
2.06

0.76

results can be obtained quickly. The PCT would rise at

‐

2 to 3 hours after infection; at 6 to 12 hours, it would

exceed the normal level; and at 24 hours, it would reach
Sepsis vs infected without sepsis

the peak. CRP usually arises at 12 to 24 hours after

Severe systematic infection vs

infection and reaches the peak after 2 to 3 days. The

without systemic infection

overall area under the SROC curve is an index that

ROC compare content

evaluates the diagnosis accuracy: the closer it is to 1, the

Sepsis vs nonsepsis

Sepsis vs nonseptic
Sepsis vs nonseptic
Sepsis vs critical ill

Sepsis vs no sepsis

better the diagnosis results. The overall area under

T A B L E 1 The basic characteristics description of included studies

the SROC between 0.5 to 0.7 represents low accuracy; a

Sepsis vs SIRS

Sepsis vs SIRS

value between 0.7 to 0.9 represents certain accuracy; a

Cross‐sectional study

2013 Cross‐sectional study

Study design
Retrospective

Cohort study
Prospective

Prospective
Kundan Kumar17 2014 Prospective

Prospective
Prospective
Longxiang Su16 2012
2014

18

Madenci20 2014
Fabian A jamies
B Jamali13 2013

Castelli14 2004
12

Massaro15 2007

Yi Yang19 2016
Hongxiang Li

Ozlem Cakir
Gian Paolo

Karin SR
Study

FIGURE 3 Forest plot—sensitivity and specificity of CRP for

the diagnosis of sepsis. CRP, C‐reactive protein
TAN ET AL. | 5

F I G U R E 4 Forest plot—DOR of CRP

for the diagnosis of sepsis. CRP,
C‐reactive protein; DOR, diagnostic odds
ratio

value above 0.9 represents high accuracy. In our study, situations. In the normal health states, the level of CRP is
the diagnostic accuracy of CRP, the overall area under usually at a low level; when there is inflammation,
the SROC curve, was 0.73 (95% CI, 0.69‐0.77), with a infection, or tissue damage in the body, CRP rises in 12 to
sensitivity and specificity of 0.80 (95% CI, 0.63‐0.90) and 18 hours, and the circulation half‐life is about 19 hours.
0.61 (95% CI, 0.50‐0.72), respectively; the diagnostic PCT is useful to identify sepsis and nonsepsis and can
accuracy of PCT, the overall area under the SROC curve, also relate well with the severity of sepsis, even
was 0.85 (95% CI, 0.82‐0.88), with a sensitivity and suggesting the prognosis condition. CRP can be used to
specificity of 0.80 (95% CI, 0.69‐0.87) and 0.77(95% CI, identify early sepsis and nonsepsis, but the level of CRP
0.60‐0.88), respectively. However, CRP is positive in has no statistical significance in the seriousness of
inflammatory condition, so it is a useful marker of the sepsis.21-25 The results showed the diagnosis accuracy
evolution of inflammatory processes; although its speci- and specificity of PCT are higher than those of CRP, and
ficity to sepsis is limited. CRP is an acute responsive both of them have a moderate degree of diagnostic value.
phase protein that is synthesized by the liver in stress In contrast with other studies, Chengfen et al26 found
that for a moderate degree of the value of PCT for the
diagnosis of sepsis in adult patients, the diagnostic
accuracy in medical patients is higher than that in
surgical patients. The overall area under the SROC curve
was 0.83 (95% CI, 0.79‐0.87), the sensitivity was 0.74 (95%
CI, 0.72‐0.76), the specificity was 0.70 (95% CI, 0.67‐0.72),

FIGURE 5 Summary receiver‐operating characteristic (SROC)

curve plot of CRP for the diagnosis of sepsis. CRP, C‐reactive
protein FIGURE 6 Publication bias of PCT. PCT, procalcitonin
6 | TAN ET AL.

FIGURE 7 Forest plot—sensitivity

and specificity of PCT for the diagnosis of
sepsis. PCT, procalcitonin

and the DOR was 10.37 (95% CI, 7.10‐15.17). Liu et al accuracy in diagnosing neonatal sepsis, regardless of
suggested that elevated PCT concentrations (the overall differences in diagnostic criteria or whether time points
area under the SROC curve being 0.77 [95% CI, 0.73‐ for testing PCT should be combined with other diagnostic
0.80], with sensitivity and specificity of 0.76 [95% CI, markers to further improve the sensitivity and accuracy
0.67‐0.82] and 0.64 [95% CI, 0.52‐0.74]) and PCT in the diagnosis of sepsis. Liu et al29 included 86 articles
nonclearance (the overall area under the SROC curve and analyzed 66 biomarkers, provided the SROC of PCT,
being 0.79 [95% CI, 0.75‐0.83], with sensitivity and CRP, interleukin 6, myeloid cells, presepsin, lipopolysac-
specificity of 0.72 [95% CI, 0.58‐0.82] and 0.77 [95% CI, charide binding protein, and CD64. Liu et al aimed to
0.55‐0.90]) are strongly associated with all‐cause mortal- evaluate the value of all the biomarkers; in our meta‐
ity in septic patients. Prkno et al27 concluded that PCT‐ analysis, we only chose the most common and repre-
guided therapy is a helpful approach to guide antibiotic sentative biomarkers of PCT and CRP and compared the
therapy and surgical interventions without a beneficial diagnosis value of PCT and CRP.
effect on mortality (hospital mortality [RR: 0.91, 95% CI, Imprecise estimates of heterogeneity can have some
0.61‐1.36], 28‐day mortality [R: 1.02, 95% CI, 1.01‐1.53]). adverse effects. They inappropriately prevent exploration
Zhangbin et al28 found that the PCT test showed good of the causes for heterogeneity with underestimation of

FIGURE 8 Forest plot—DOR of PCT

for the diagnosis of sepsis. DOR, diagnostic
odds ratio; PCT, procalcitonin
TAN ET AL. | 7

and infection without sepsis were other main factors

which might account for the heterogeneity. Because I2 of
Figures 7 and 8 was greater than 90%, we also need to
conduct a sensitivity analysis to explore the heterogeneity.

10 | C ON C LU S I O N S

Our results together indicate a moderate degree of value

of PCT and CRP for diagnosis of sepsis in adult patients.
However, the accuracy of the diagnosis of PCT is higher
than that of CRP based on the results of the overall area
under the SROC curve. Further studies are needed to
define the optimal cutoff point (PCT, CRP) and the
diagnostic indexes in different disease stages.

ACKN OWLEDGMENT
FIGURE 9 Summary receiver‐operating characteristic (SROC)
The authors would like to thank Shanghai Baoshan
curve plot of PCT for the diagnosis of sepsis. PCT, procalcitonin
Traditional Chinese Medicine—Integrated Hospital.

heterogeneity. Therefore, overestimation of heterogeneity

could prevent a meta‐analysis from being actually done, ETHICAL AP PROVAL
which causes an inappropriate exploration of the cause.
In our study, the I2 of Figures 7 and 8 was greater than The Research Ethics Committee of Shanghai Baoshan
90%. This I2 estimates might prompt us to exhaust all Traditional Chinese Medicine—integrated Hospital ap-
possibilities of subgroup analyses. proved the collection of tissue samples for research.
However, there are some limitations of this analysis
that should be paid attention to. The limitations are as
ORCID
follows: (1) prospective study, retrospective study, cross‐
sectional study, and cohort study were included;(2) Liandong Zhang http://orcid.org/0000-0002-3570-6308
differences in the inclusion criteria and exclusion criteria
for patients in different articles; (3) different patients with
previous disease and treatments were unavailable; (4) all RE FER E NCES
the included studies were English or Chinese articles; (5) 1. Czura CJ. “Merinoff Symposium 2010: sepsis”—speaking with
pooled data were used for analysis, and individual one voice. Mol Med. 2011;17:2‐3.
patients’ data were unavailable, so it limited us from 2. Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and
making a more comprehensive analysis; (6) the different organ failure and guidelines for the use of innovative therapies
cutoff values of included studies; (7) the difference group in sepsis. The ACCP/SCCM Consensus Conference Committee.
standard and ROC compare content; (8) the sampling American College of Chest Physicians/Society of Critical Care
Medicine. Chest. 1992;101(6):1644‐1655.
time of PCT or CRP were different.
3. Dombrovskiy VY, Martin AA, Sunderram J, Paz HL. Rapid
In addition, there was significant heterogeneity be- increase inhospitalization and mortality rates for severe sepsis
tween the included studies for pooled estimate diagnostic in the United States: a trend analysis from 1993 to 2003. Crit
accuracy of CRP and PCT, which might be caused by the Care Med. 2007;35:1244‐1250.
different sepsis patient population and state of illness. For 4. Iskander KN, Osuchowski MF, Stearns‐Kurosawa DJ, et al.
example, in Mohsen et als’ study, the sepsis patients with Sepsis: multiple abnormalities, heterogeneous responses, and
febrile neutropenia were studied, while sepsis in alcoholic evolving understanding. Physiol Rev. 2013;93:1247‐1288.
5. Garnacho‐Montero J, Ortiz‐Leyba C, Herrera‐Melero I, et al.
hepatitis was studied in Kundan et als’ study. Moreover,
Mortality and morbidity attributable to inadequate empirical
because of the different state of illness, the cutoff values of
antimicrobial therapy in patients admitted to the ICU with
PCT or CRP were various between each study, which may sepsis: a matched cohort study. J Antimicrob Chemother.
lead to a great heterogeneity for estimating the diagnostic 2008;61(2):436‐441.
accuracy. Furthermore, the different controls like non- 6. Parrillo JE, Parker MM, Natanson C, et al. Septic shock in
sepsis, critical illness, the absence of systemic infection humans. Advances in the understanding of pathogenesis,
8 | TAN ET AL.

cardiovascular dysfunction, and therapy. Ann Intern Med. 19. Yang Y, Xie J, Guo F, et al. Combination of C‐reactive protein,
1990;113(3):227‐242. procalcitonin and sepsis‐related organ failure score for the
7. Bates DW. Predicting bacteremia in hospitalized patients. A diagnosis of sepsis in critical patients. Ann Intensive Care.
prospectively validated model. Ann Intern Med. 1990;113(7): 2016;6:51.
495‐500. 20. Çakır madenci Ö, Yakupoğlu S, Benzonana N, Yücel N, Akbaba
8. Bates DW, Sands K, Miller E, et al. Predicting bacteremia in D, Orçun Kaptanağası A. Evaluation of soluble CD 14 subtype
patients with sepsis syndrome. Academic Medical Center (presepsin) in burn sepsis. Burns. 2014;40:664‐669.
Consortium Sepsis Project Working Group. J Infect Dis. 21. Ríostoro JJ, Márquezcoello M, García‐Álvarez JM, et al. Soluble
1997;176(6):1538‐1551. membrane receptors, interleukin 6, procalcitonin and C
9. Pittet D, Rangel‐Frausto S, Li N, et al. Systemic inflammatory reactive protein as prognostic markers in patients with severe
response syndrome, sepsis, severe sepsis and septic shock: sepsis and septic shock. PLoS One. 2017;12(4):e0175254.
incidence, morbidities and outcomes in surgical ICU patients. 22. Hatherill M, Tibby SM, Sykes K, Turner C, Murdoch IA.
Intensive Care Med. 1995;21(4):302‐309. Diagnostic markers of infection: comparison of procalcitonin
10. Moyer MW. New biomarkers sought for improving sepsis with C reactive protein and leucocyte count. Arch Dis Child.
management and care. Nat Med. 2012;18(7):999‐999. 1999;81:417‐421.
11. Uzzan B, Cohen R, Nicolas P, Cucherat M, Perret GY. 23. Pepys MB, Hirschfield GM. C‐reactive protein: a critical update.
Procalcitonin as a diagnostic test for sepsis in critically ill J Clin Invest. 2003;111:1805‐1812.
adults and after surgery or trauma: a systematic review and 24. Luzzani A, Polati E, Dorizzi R, Rungatscher A, Pavan R,
meta‐analysis. Crit Care Med. 2006;34(7):1996‐2003. Merlini A. Comparison of procalcitonin and C‐reactive protein
12. Li HX, Liu ZM, Zhao SJ, Zhang D, Wang SJ, Wang YS. as marker of sepsis. Crit Care Med. 2003;31:1737‐11741.
Measuring both procalcitonin and C‐reactive protein for a 25. Jiang X‐G, Wang X‐B, Wang R‐S. The clinical significance of
diagnosis of sepsis in critically ill patients. J Int Med Res. serum procalcitonin and C‐reactive protein in sepsis. Chin J
2014;42(4):1050‐1059. Respir Crit Care Med. 2009;8(5):429‐431.
13. Jamali B, Khorvash F, Meidani M, Abolghasemi H. Procalci- 26. Chengfen Y, Tong L, Xinjing G, Zhibo L, Lei X. Accuracy of
tonin and quantitative C‐reactive protein role in the early procalcitonin for diagnosis of sepsis in adults: a meta‐analysis.
diagnosis of sepsis in patients with febrile neutropenia. South Chin Crit Care Med. 2015;27(9):743‐749.
Asian J Cancer. 2013;2(4):216‐219. 27. Liu D, Su L, Han G, Yan P, Xie L. Prognostic value of
14. Castelli G, Pognani C, Meisner M, Stuani A, Bellomi D, Sgarbi procalcitonin in adult patients with sepsis: a systematic review
L. Procalcitonin and C‐reactive protein during systemic and meta‐analysis. PLoS One. 2015;10:e0129450.
inflammatory response syndrome, sepsis and organ dysfunc- 28. Prkno A, Wacker C, Brunkhorst FM, Schlattmann P. Procalci-
tion. Crit Care. 2004;8(4):R234‐R242. tonin‐guide therapy in intensive care unit patients wit severe
15. Massaro KS, Costa SF, Leone C, Chamone D. Procalcitonin and sepsis and septic shock‐a systematic review and meta‐analysis.
C‐reactive protein as severe systemic infection markers in Crit Care. 2013;177:R291.
febrile neutropenic adults. BMC Infect Dis. 2007;7:137. 29. Liu Y, Hou J, Li Q, Chen Kj, Wang SN, Wang Jm. Biomarkers
16. Su L, Han B, Liu C, et al. Value of soluble TREM‐1, for diagnosis of sepsis in patients with systemic inflammatory
procalcitonin, and C‐reactive protein serum levels as biomar- response syndrome: a systematic review and meta‐analysis.
kers for detecting bacteremia among sepsis patients with new Springer Plus. 2016;5:2091.
fever in intensive care units: a prospective cohort study. BMC
Infect Dis. 2012;12:157.
17. Kumar K, Mohindra S, Raj M, Choudhuri G. Procalcitonin as a
How to cite this article: Tan M, Lu Y, Jiang H,
marker of sepsis in alcoholic hepatitis. Hepatol Int. 2014;8:
Zhang L. The diagnostic accuracy of procalcitonin
436‐442.
18. Jaimes FA, De la rosa GD, Valencia ML, et al. A latent class and C‐reactive protein for sepsis: A systematic
approach for sepsis diagnosis supports use of procalcitonin in review and meta‐analysis. J Cell Biochem. 2018;1‐8.
the emergency room for diagnosis of severe sepsis. BMC https://doi.org/10.1002/jcb.27870
Anesthesiol. 2013;13:23.