Accepted Manuscript

Prognostic value of plasma brain natriuretic peptide value for

patients with sepsis: A meta-analysis

Yu-long Bai, Bang-Li Hu, Han-Chun Wen, Yi-Ling Zhang, Ji-Jin

Zhu

PII: S0883-9441(18)30399-X
DOI: doi:10.1016/j.jcrc.2018.08.040
Reference: YJCRC 53032
To appear in: Journal of Critical Care

Please cite this article as: Yu-long Bai, Bang-Li Hu, Han-Chun Wen, Yi-Ling Zhang, Ji-Jin
Zhu , Prognostic value of plasma brain natriuretic peptide value for patients with sepsis: A
meta-analysis. Yjcrc (2018), doi:10.1016/j.jcrc.2018.08.040

This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
 ACCEPTED MANUSCRIPT

Prognostic value of plasma brain natriuretic peptide value for patients with sepsis: a

meta-analysis

Yu-long Bai1, Bang-Li Hu2, Han-Chun Wen3, Yi-Ling Zhang3, Ji-Jin Zhu1*

1 Department of Emergency, First Affiliated Hospital of Guangxi Medical University,

PT
Shuangyong Road 6, Nanning 530021, Guangxi, PR China;

RI
2 Department of Research, Affiliated Tumor Hospital of Guangxi Medical University, Hedi

SC
Road 71, Nanning 530021, Guangxi, PR China;

3 Department of Critical Care Medicine, First Affiliated Hospital of Guangxi Medical

NU
University, Shuangyong Road 6, Nanning 530021, Guangxi, PR China;
MA

*Correspondence to: Ji-Jin Zhu

Department of Emergency,
D

First Affiliated Hospital of Guangxi Medical University,

E

Shuangyong Road 6, Nanning 530021, Guangxi, PR China.

PT

Tel: +86-0771-5335022; Fax +86-0771-5335022

CE

Email: 2637874438@qq.com
AC

Yu-long Bai: baiyulong_1983@aliyun.com

Bang-Li Hu: gxhubl@163.com

Han-Chun Wen: 1037315348@qq.com

Yi-Ling Zhang: 1458115645@qq.com

Ji-Jin Zhu: 2637874438@qq.com

 ACCEPTED MANUSCRIPT

Abstract

Purpose

The aim of this meta-analysis was to clarify the diagnostic role of plasma BNP and

NT-proBNP in predicting mortality for septic patients.

Methods

A systematic review was conducted prior to January 2018. Summary sensitivity, specificity,

PT
positive likelihood ratio (PLR), negative likelihood ratio (NLR) and diagnostic odds ratio

RI
(DOR) of the prognostic value of plasma BNP and NT-proBNP for septic patients. The area

SC
under the receiver operating curves (AUROC) were used to summarize overall test

performance.
NU
Results

Twenty-two studies with 3417 septic patients were selected in the analysis. The summary
MA

sensitivity, specificity, PLR, NLR, DOR and the AUROC of the overall analysis of BNP

were: 0.84, 0.73, 3.1, 0.22, 14, 0.85; and these values of NT-proBNP were: 0.71, 0.73, 2.6,
D

0.39, 7 and 0.7 respectively; Subgroup analysis and meta-regression analyses showed that the
E

tested method and observation endpoint influenced the summary sensitivity, specificity of
PT

BNP, but the tested day, tested method or observation endpoint did not influence the
CE

summary sensitivity, specificity of NT-proBNP.

Conclusions
AC

This meta-analysis indicates that both elevated plasma BNP and NT-proBNP have moderate

predicts value for the mortality of septic patients, and the tested method and observation

endpoint influence the results of BNP.

Key words: Sepsis; Septic shock; Prognosis; Meta-analysis

 ACCEPTED MANUSCRIPT

Introduction

Sepsis and septic shock have become increasingly important clinical problems. A total of

PT
40% of patients with sepsis have myocardial dysfunction, and the mortality rate of patients

RI
with sepsis with cardiac dysfunction is as high as 70% [1]. Echocardiography can accurately

SC
detect myocardial depression, and the bedside echocardiograms have become an increasing

trend; however, the diagnosis of myocardial conditions via echocardiography is subject to

NU
operator experience [2]. Many studies have identified several biomarkers associated with the

prognosis of patients with sepsis, but these biomarkers have limitations and their prognostic
MA

value remains poor [3-5]. Therefore, finding a reliable biomarker to predict the prognosis of

patients with sepsis is crucial in improving survival.

D

Brain natriuretic peptide (BNP) is a prohormone secreted by cardiomyocytes. N-terminal

E

pro-BNP (NT-proBNP) is an inactive metabolite of BNP that has a markedly longer plasma
PT

half-life and is more stable than BNP, making it more applicable for clinical use [6].
CE

Compared with echocardiography, plasma BNP or NT-proBNP evaluation is simple,

inexpensive, and widely available. Increased plasma levels of both BNP and proBNP have
AC

been identified as early markers of myocardial dysfunction and increased mortality in patients

admitted to the intensive care unit (ICU) [7, 8]. Elevated levels of plasma BNP and

NT-proBNP have also been shown to be markers of unfavorable prognosis in patients with

severe sepsis and septic shock [9-11].

Some studies reported the prognostic value of plasma BNP and NT-proBNP for the

patients with sepsis, but the results substantially varied generally because of difference in the
 ACCEPTED MANUSCRIPT

sample size and testing methods for the markers. A meta-analysis uses a statistical approach

to combine the results from multiple studies and provides more reliable results than

individual studies. Previously, a meta-analysis was performed to explore the prognostic value

of BNP in patients with sepsis [12]; however, only a small number of studies were included

and the potential confounders that might affect the prognostic value of BNP were not

investigated. Thus, the generalizability of the results was limited. In this study, we aimed to

PT
examine the role of plasma BNP and NT-proBNP in predicting mortality in patients with

RI
sepsis.

SC
Materials and Methods
NU
Search strategy

This meta-analysis was performed in accordance with the guidelines of the Preferred
MA

Reporting Items for Systematic Review and Meta-analyses. The electronic databases PubMed,

Cochrane Library, Web of Science, and Chinese National Knowledge Infrastructure were
D

comprehensively searched for eligible articles published before January 2018 using the
E

following search terms: “septic shock” or “Shock, Septic,” “Natriuretic Peptide, Brain” or
PT

“BNP,” “N-terminal pro-B-type” or “pro-BNP,” and “prognostic” or “prognosis.” The search

CE

strategy was a combination of at least three terms, for example, “septic shock” combined with

“BNP” and “prognostic.” The included articles were limited to human studies but not limited
AC

by language. Conference abstracts were not selected because they generally do not contain

sufficient data. Relevant articles were also searched using the related articles function in

PubMed. References within the identified articles were also searched manually.

Inclusion and exclusion criteria

 ACCEPTED MANUSCRIPT

The inclusion criteria were as follows: (1) the diagnosis of sepsis and septic shock was in

accordance with 2008 guidelines [13] or the updated 2013 [14] or 2016 version [15] for

sepsis and septic shock, (2) the patients were given standard treatment for sepsis; (3) the

study evaluated the prognostic value of plasma BNP or NT-proBNP for the patients with

sepsis; (4) data on the sensitivity and specificity of BNP or NT-proBNP for sepsis were

provided; (5) the patient outcome (survival and non-survival) and study endpoints were

PT
provided. Meanwhile, studies were excluded if they were review articles, letters, case reports,

RI
studies on cell lines, animal experiments, and human xenografts. If the same patient cohort

SC
was reported in several articles, we chose the most recent study.
NU
Data extraction and quality assessment

The following data were extracted: the name of the first author, year of publication, study
MA

location (country), cut-off of plasma BNP or NT-proBNP, number of patients who survived

and died, sensitivity and specificity value of BNP or NT-proBNP, study design, method of
D

evaluating plasma BNP or NT-proBNP, and endpoint of observation. The quality of the
E

studies was assessed using Quality Assessment of Diagnostic Accuracy Studies (QUADAS)
PT

[16], which is an evidence-based quality assessment tool developed for systematic reviews of
CE

studies of diagnostic accuracy. The maximum QUADAS score is 14, with a 6 score

indicating high quality. Two reviewers independently evaluated the eligibility of the studies.
AC

Disagreements were resolved by consensus.

Statistical analysis

The following summary measures were calculated for each study: sensitivity and

specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), and diagnostic

odds ratio (DOR) with the corresponding 95% confidence intervals (CIs). The Cochran Q test
 ACCEPTED MANUSCRIPT

and the inconsistency index (I2) were used to assess the threshold effect as an important

component of the source of variation of the studies. In general, I2 values <25% indicate mild

heterogeneity; 25%-50%, moderate heterogeneity; and >50%, significant heterogeneity. A

fixed-effects model (Mantel-Haenszel method) was used to calculate parameters when there

was low heterogeneity; otherwise, a random-effects model (DerSimonian and Laird) was

used. Subgroup analysis and meta-regression analyses were conducted to explore and explain

PT
the diversity (heterogeneity) among the results of different studies. Publication bias was

RI
assessed using Deek’s test [17]. All statistical tests were performed using STATA 11.2

SC
software (Stata Corp, College Station, TX) with 2-tailed P values. A P value <0.05 was

considered significant.
NU
Results
MA

Study selection process

Initially, 106 articles were retrieved. After screening the abstracts, 46 articles were
D

excluded because they were either reviews, animal studies, case reports, or irrelevant to the
E

current meta-analysis. Of the remaining 60 articles, 38 were further excluded after screening
PT

their full-text because they did not provide data on the sensitivity and specificity of BNP or
CE

NT-proBNP (n=18) or data on the survival rate of patients (n=15), and the prognostic hazard

ratio of BNP or NT-proBNP was provided instead of the sensitivity and specificity values
AC

(n=5). Finally, 22 studies [18-39] with 3417 patients with sepsis were included in our

meta-analysis. A flow chart of the study selection process is shown in Fig. 1.

Characteristics of included studies

The characteristics of the included studies are shown in Table 1. All the patients with

sepsis meet the criteria for severe sepsis or septic shock and were given standard treatment.
 ACCEPTED MANUSCRIPT

BNP or pro-BNP were evaluated via electrochemiluminescence immunoassay using Elecsys

2010 analyzer (Roche), Biosite Triage immunoassay, or enzyme-linked immunosorbent assay.

Most of the studies were prospective in design and used 28-day mortality as the endpoint.

Other endpoints included hospital mortality and ICU mortality. The quality of the studies was

high, with QUADAS scores ranging from 12 to 14.

PT
Prognostic value of plasma BNP for patients with sepsis

RI
Fourteen studies that included 2380 patients with sepsis evaluated the prognostic value of

SC
plasma BNP. The combined data showed that plasma BNP has moderate sensitivity and

moderate specificity, with pooled sensitivity and specificity of 0.84 (95% CI: 0.77-0.89) and
NU
0.73 (95% CI: 0.65-0.79), respectively. The PLR, NLR, and DOR was 3.1 (95% CI: 2.3-4.1),

0.22 (95% CI: 0.15-0.34), and 14 (95% CI: 7-27), respectively. However, significant
MA

heterogeneity was observed among the studies regarding their combined sensitivity

(I2=96.07%, P<0.01) and specificity (I2=74.21%, P<0.01) (Fig. 2). The area under the curve
D

(AUC) of the summary receiver operating characteristic (SROC) curve was 0.85 (0.81-0.88),
E

suggesting a moderate diagnostic accuracy (Fig. 3). No evidence of publication bias was
PT

observed across the studies (Deek’s test: P=0.112; Fig. 4).

CE

Subgroup analysis and meta-regression analyses for plasma BNP

AC

The subgroup analysis and meta-regression analyses for the prognostic value of plasma

BNP was based on the testing day, testing method, or observation endpoint. The Biosite

Triage immunoassay has higher summary specificity than immunoradiometric assay, and the

sensitivity and specificity of NT-proBNP was lower in the 28-day mortality endpoint (Table

3). These results indicate that the testing method or observation endpoint may have caused

the heterogeneity of the studies.

 ACCEPTED MANUSCRIPT

Prognostic value of NT-proBNP for patients with sepsis

Nine studies with 691 patients with sepsis assessed the prognostic value of plasma

NT-proBNP for sepsis. The pooled results showed that NT-proBNP has a moderate

prognostic value, with the pooled sensitivity and specificity of 0.71 (95% CI: 0.60-0.80) and

0.73 (95% CI: 0.63-0.81), respectively. The PLR, NLR, and DOR was 2.6 (95% CI: 1.9-3.6),

PT
0.39 (95% CI: 0.28-0.55), and 7 (95% CI: 4-12), respectively. However, there was significant

RI
heterogeneity across the studies in the pooled sensitivity (I2=88.86%, P<0.01) and specificity

analysis (I2=67.25%, P<0.01) (Fig. 5.) The AUC of SROC was 0.78 (0.75-0.82), suggesting a

SC
moderate diagnostic accuracy (Fig. 6). The Deek’s test indicated no significant publication
NU
bias across the included studies (P=0.191; Fig. 7).
MA

Subgroup analysis and meta-regression analyses for plasma NT-proBNP

Because there was significant heterogeneity in the summary sensitivity and specificity,
D

we further explored the potential causes of the heterogeneity in the meta-analysis using
E

subgroup analysis and meta-regression analyses based on the testing day, testing method, or
PT

observation endpoint. The results showed that the summary sensitivity and specificity of
CE

plasma NT-proBNP tested at day one is not significantly different compared with that tested

in other days. Similar results were also found in the subgroup analysis for testing the method
AC

or observation endpoint (Table 2).

Discussion

In this study, we evaluated the value BNP or NT-proBNP in predicting the mortality of

patients with sepsis. Our meta-analysis of twenty-two studies showed that both BNP and

NT-proBNP have moderate prognostic value for patients with sepsis. Because there was
 ACCEPTED MANUSCRIPT

heterogeneity across the studies in terms of combined sensitivity and specificity, we analyzed

the prognostic value of these biomarkers using subgroup analysis of the different testing time,

testing method, or observation endpoint, which yielded more important information than

previous studies. We also performed a sensitivity analysis by alternately removing studies

that may have been the source of heterogeneity, and the results were similar to that of the

main analysis, indicating the reliability of the results.

PT
Sepsis and septic shock are severe systemic inflammatory response to infection.

RI
Myocardial dysfunction induces the endogenous release of various mediators with deleterious

SC
cardiovascular effects; thus, it is a crucial complication of septic shock [40]. BNP and

NT-proBNP are cardiac hormones with diuretic, natriuretic and vasorelaxation properties and
NU
play fundamental roles in regulating cardiac filling pressure and intravascular volume

homeostasis [41, 42]. As such, plasma BNP and NT-proBNP are widely used as a valuable
MA

biomarker of left ventricular (LV) dysfunction and increased LV filling pressure [43, 44].

However, increased plasma levels of BNP and NT-proBNP are not specific for heart failure,
D

and circulating levels may be influenced by several cardiac and non-cardiac conditions that
E

cause myocardial strain and affect the clearance of these prohormones, including myocardial
PT

ischemia, cardiac arrhythmias, sepsis, shock, and anemia.

CE

Plasma BNP and NT-proBNP levels predict the prognosis in patients with severe sepsis

and septic shock. However, the association between these peptides and cardiac filling
AC

pressures in sepsis and septic shock is unclear. Previously, Wang et al [12] performed a

meta-analysis of the prognostic value of BNP or NT-proBNP for sepsis using 12 studies and

found that elevated BNP or NT-proBNP were significantly associated with increased risk of

mortality of patients with sepsis. The pooled sensitivity and specificity for BNP were 79%

(95% CI: 75-83) and 60% (95% CI: 57 -62). However, they did not analyze the prognostic

value of BNP or NT-proBNP separately; Thus, their individual values cannot be determined.
 ACCEPTED MANUSCRIPT

In addition, they did not separate the mortality endpoint, which included 28-day, ICU, and

in-hospital mortality. Thus, the conclusions are vague and could not reflect the true

prognostic value of BNP or NT-proBNP. A major strength of our study is that the sample size

was larger than the ideal one. Meanwhile, there were also some limitations that need to be

noted in the present study. First, because the optimal cut-off point of BNP or NT-proBNP

substantially varied across the studies, we only performed subgroup analysis by dividing the

PT
studies into two groups. Second, although the testing day and the endpoint differed in the

RI
studies, we only divided the studies into two groups in the meta-regression analysis. Further

SC
analysis should be performed on the specific variables in the future. Third, the data used in

the present study were not adjusted according to the potential confounders, such as the
NU
patients’ age and sex. It should be noted that the pooled estimates reflected unadjusted

associations between BNPs and all-cause mortality. Fourth, the cause of sepsis varied and
MA

included lung, intra-abdominal, and urinary tract infection, which might influence the

robustness of the results. Collectively, the differences in lead time bias, endpoint measures,
D

control for cardiac function, and sampling time might influence the prognostic value of BNP
E

or NT-proBNP in patients with sepsis. Our findings should be validated in future studies that
PT

address these limitations.

CE

Conclusions
AC

This meta-analysis showed that elevated plasma BNP or NT-proBNP appears to be a

marker of outcome in sepsis.. However, large prospective studies that consider potential

confounders are needed to accurately determine the prognostic value of plasma BNP or

NT-proBNP in these patients. Considering the inconsistent results on the prognostic value of

BNP or NT-proBNP for patients with sepsis, our results provide clinicians with information

to evaluate the importance of BNP levels on determining patient prognosis.

 ACCEPTED MANUSCRIPT

Conflict of Interest

The authors declare that they have no conflict of interest.

Funding

This research did not receive any specific grant from funding agencies in the public,

PT
commercial, or not-for-profit sectors.

RI
SC
Ethical approval

This article does not contain any studies with human participants or animals performed by
NU
any of the authors.
MA
E D
PT
CE
AC
 ACCEPTED MANUSCRIPT

Reference

[1] Li Y, Ge S, Peng Y, Chen X. Inflammation and cardiac dysfunction during sepsis,

muscular dystrophy, and myocarditis. Burns Trauma 2013;1(3):109-21.

[2] Li Z, Zhang E, Hu Y, Liu Y, Chen B. High Serum sTREM-1 Correlates With

Myocardial Dysfunction and Predicts Prognosis in Septic Patients. Am J Med Sci

PT
2016;351(6):555-62.

RI
[3] Hamed S, Behnes M, Pauly D, Lepiorz D, Barre M, Becher T, et al. Diagnostic value

SC
of Pentraxin-3 in patients with sepsis and septic shock in accordance with latest

sepsis-3 definitions. BMC Infect Dis 2017;17(1):554.

NU
[4] Kosir M, Podbregar M. Advances in the Diagnosis of Sepsis: Hydrogen Sulfide as a

Prognostic Marker of Septic Shock Severity. EJIFCC 2017;28(2):134-41.

MA

[5] Wang B, Li J, Gao HM, Xing YH, Lin Z, Li HJ, et al. Necroptosis regulated proteins

expression is an early prognostic biomarker in patient with sepsis: a prospective

D

observational study. Oncotarget 2017;8(48):84066-73.

E

[6] Pfister R, Scholz M, Wielckens K, Erdmann E, Schneider CA. Use of NT-proBNP in

PT

routine testing and comparison to BNP. Eur J Heart Fail 2004;6(3):289-93.

CE

[7] Kotanidou A, Karsaliakos P, Tzanela M, Mavrou I, Kopterides P, Papadomichelakis E,

et al. Prognostic importance of increased plasma amino-terminal pro-brain natriuretic

AC

peptide levels in a large noncardiac, general intensive care unit population. Shock

2009;31(4):342-7.

[8] Meyer B, Huelsmann M, Wexberg P, Delle Karth G, Berger R, Moertl D, et al.

N-terminal pro-B-type natriuretic peptide is an independent predictor of outcome in

an unselected cohort of critically ill patients. Crit Care Med 2007;35(10):2268-73.

 ACCEPTED MANUSCRIPT

[9] Khoury J, Arow M, Elias A, Makhoul BF, Berger G, Kaplan M, et al. The prognostic

value of brain natriuretic peptide (BNP) in non-cardiac patients with sepsis, ultra-long

follow-up. J Crit Care 2017;42:117-22.

[10] Masson S, Caironi P, Fanizza C, Carrer S, Caricato A, Fassini P, et al. Sequential

N-Terminal Pro-B-Type Natriuretic Peptide and High-Sensitivity Cardiac Troponin

Measurements During Albumin Replacement in Patients With Severe Sepsis or Septic

PT
Shock. Crit Care Med 2016;44(4):707-16.

RI
[11] Singh H, Ramai D, Patel H, Iskandir M, Sachdev S, Rai R, et al. B-Type Natriuretic

SC
Peptide: A Predictor for Mortality, Intensive Care Unit Length of Stay, and Hospital

Length of Stay in Patients With Resolving Sepsis. Cardiol Res 2017;8(6):271-5.

NU
[12] Wang F, Wu Y, Tang L, Zhu W, Chen F, Xu T, et al. Brain natriuretic peptide for

prediction of mortality in patients with sepsis: a systematic review and meta-analysis.

MA

Crit Care 2012;16(3):R74.

[13] Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R, et al. Surviving
D

Sepsis Campaign: international guidelines for management of severe sepsis and septic
E

shock: 2008. Crit Care Med 2008;36(1):296-327.

PT

[14] Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, et al. Surviving
CE

Sepsis Campaign: international guidelines for management of severe sepsis and septic

shock, 2012. Intensive Care Med 2013;39(2):165-228.

AC

[15] Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al.

The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3).

JAMA 2016;315(8):801-10.

[16] Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J. The development of

QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included

in systematic reviews. BMC Med Res Methodol 2003;3:25.

 ACCEPTED MANUSCRIPT

[17] Deeks JJ, Macaskill P, Irwig L. The performance of tests of publication bias and other

sample size effects in systematic reviews of diagnostic test accuracy was assessed. J

Clin Epidemiol 2005;58(9):882-93.

[18] Brueckmann M, Huhle G, Lang S, Haase KK, Bertsch T, Weiss C, et al. Prognostic

value of plasma N-terminal pro-brain natriuretic peptide in patients with severe sepsis.

Circulation 2005;112(4):527-34.

PT
[19] Charpentier J, Luyt CE, Fulla Y, Vinsonneau C, Cariou A, Grabar S, et al. Brain

RI
natriuretic peptide: A marker of myocardial dysfunction and prognosis during severe

SC
sepsis. Crit Care Med 2004;32(3):660-5.

[20] Mokart D, Sannini A, Brun JP, Faucher M, Blaise D, Blache JL, et al. N-terminal
NU
pro-brain natriuretic peptide as an early prognostic factor in cancer patients

developing septic shock. Crit Care 2007;11(2):R37.

MA

[21] Papanikolaou J, Makris D, Mpaka M, Palli E, Zygoulis P, Zakynthinos E. New

insights into the mechanisms involved in B-type natriuretic peptide elevation and its
D

prognostic value in septic patients. Crit Care 2014;18(3):R94.

E

[22] Post F, Weilemann LS, Messow CM, Sinning C, Munzel T. B-type natriuretic peptide
PT

as a marker for sepsis-induced myocardial depression in intensive care patients. Crit

CE

Care Med 2008;36(11):3030-7.

[23] Qian A, Zhang M, Zhao G. Dynamic detection of N-terminal pro-B-type natriuretic

AC

peptide helps to predict the outcome of patients with major trauma. Eur J Trauma

Emerg Surg 2015;41(1):57-64.

[24] Rivers EP, McCord J, Otero R, Jacobsen G, Loomba M. Clinical utility of B-type

natriuretic peptide in early severe sepsis and septic shock. J Intensive Care Med

2007;22(6):363-73.
 ACCEPTED MANUSCRIPT

[25] Roch A, Allardet-Servent J, Michelet P, Oddoze C, Forel JM, Barrau K, et al. NH2

terminal pro-brain natriuretic peptide plasma level as an early marker of prognosis

and cardiac dysfunction in septic shock patients. Crit Care Med 2005;33(5):1001-7.

[26] Sturgess DJ, Marwick TH, Joyce C, Jenkins C, Jones M, Masci P, et al. Prediction of

hospital outcome in septic shock: a prospective comparison of tissue Doppler and

cardiac biomarkers. Crit Care 2010;14(2):R44.

PT
[27] Ueda S, Nishio K, Akai Y, Fukushima H, Ueyama T, Kawai Y, et al. Prognostic value

RI
of increased plasma levels of brain natriuretic peptide in patients with septic shock.

SC
Shock 2006;26(2):134-9.

[28] Varpula M, Pulkki K, Karlsson S, Ruokonen E, Pettila V. Predictive value of

NU
N-terminal pro-brain natriuretic peptide in severe sepsis and septic shock. Crit Care

Med 2007;35(5):1277-83.
MA

[29] Yucel T, Memis D, Karamanlioglu B, Sut N, Yuksel M. The prognostic value of atrial

and brain natriuretic peptides, troponin I and C-reactive protein in patients with sepsis.
D

Exp Clin Cardiol 2008;13(4):183-8.

E

[30] Zhao HY, An YZ, Liu F. [Prognostic values of B-type natriuretic peptide in severe
PT

sepsis and septic shock]. Zhongguo Wei Zhong Bing Ji Jiu Yi Xue 2009;21(5):293-5.
CE

[31] Chen Y, Li C. Prognostic significance of brain natriuretic peptide obtained in the ED

in patients with SIRS or sepsis. Am J Emerg Med 2009;27(6):701-6.

AC

[32] Hu BC, Wang YJ, Ge WD, Li FZ, Sun RH. [Combination of B-type brain natriuretic

peptide with left ventricular diastolic dysfunction for prediction of mortality in the

patients with septic shock]. Zhonghua Yi Xue Za Zhi 2016;96(29):2295-300.

[33] Park BH, Park MS, Kim YS, Kim SK, Kang YA, Jung JY, et al. Prognostic utility of

changes in N-terminal pro-brain natriuretic Peptide combined with sequential organ

 ACCEPTED MANUSCRIPT

failure assessment scores in patients with acute lung injury/acute respiratory distress

syndrome concomitant with septic shock. Shock 2011;36(2):109-14.

[34] Li SW, Sun FH, Tao ZX, Song J, Zhao YF. Distribution of pathogens isolated from

patients with acute abdomen patients complicated with septic shock and value of brain

natriuretic peptide in assessment of prognosis. Chinese Journal of Nosocomiology

2017;27(16):3707-310.

PT
[35] Li ZH, Xiao J, Li JZ. Prognostic value of N-terminal pro-B-type natriuretic peptide in

RI
critically ill patients. hinese Critical Care Medicine 2011;23(3):179-82.

SC
[36] Xing B, Wang XZ, Zeng Y, Tan SF, Huang S. Prognostic value of plasma B-type

natriuretic peptide in patients with septic shock. China Journal of Modern Medicine
NU
2013;23(32):75-9.

[37] Yang Y, Wang XP, Chen XJ, Ren H. Prognostic value of brain natriuretic peptides
MA

in patients with septic shock. Guangdong Medical Journal 2016;37(17):2631-3.

[38] Phua J, Koay ES, Lee KH. Lactate, procalcitonin, and amino-terminal pro-B-type
D

natriuretic peptide versus cytokine measurements and clinical severity scores for
E

prognostication in septic shock. Shock 2008;29(3):328-33.

PT

[39] Perman SM, Chang AM, Hollander JE, Gaieski DF, Trzeciak S, Birkhahn R, et al.
CE

Relationship between B-type natriuretic peptide and adverse outcome in patients with

clinical evidence of sepsis presenting to the emergency department. Acad Emerg Med
AC

2011;18(2):219-22.

[40] Martin GS. Sepsis, severe sepsis and septic shock: changes in incidence, pathogens

and outcomes. Expert Rev Anti Infect Ther 2012;10(6):701-6.

[41] Levin ER, Gardner DG, Samson WK. Natriuretic peptides. N Engl J Med

1998;339(5):321-8.
 ACCEPTED MANUSCRIPT

[42] Tschope C, Kasner M, Westermann D, Walther T, Gaub R, Poller WC, et al. Elevated

NT-ProBNP levels in patients with increased left ventricular filling pressure during

exercise despite preserved systolic function. J Card Fail 2005;11(5 Suppl):S28-33.

[43] Park SJ, Cho EJ, Ahn J, Carriere K, Kim EK, Lee GY, et al. Additive prognostic

values of NT-proBNP and exercise stress echocardiography in asymptomatic patients

with degenerative mitral regurgitation and preserved left ventricular ejection fraction.

PT
Int J Cardiol 2017;236:387-92.

RI
[44] Siva Sankara C, Rajasekhar D, Vanajakshamma V, Praveen Kumar BS, Vamsidhar A.

SC
Prognostic significance of NT-proBNP, 3D LA volume and LV dyssynchrony in

patients with acute STEMI undergoing primary percutaneous intervention. Indian

NU
Heart J 2015;67(4):318-27.
MA
E D
PT
CE
AC
 ACCEPTED MANUSCRIPT

Figure legend

Figure 1 Flow diagram of the study selection process.

Figure 2 Summary sensitivity, specificity plotted on forest graphs for BNP in predicting the

mortality of septic patients.

Figure 3 SROC curve graph for BNP in predicting the mortality of septic patients.

Figure 4 Deek’s funnel plot asymmetry test of BNP in predicting the mortality of septic

PT
patients.

RI
Figure 5 Summary sensitivity, specificity plotted on forest graphs for NT-proBNP in

SC
predicting the mortality of septic patients.

Figure 6 SROC curve graph for NT-proBNP in predicting the mortality of septic patients.
NU
Figure 7 Deek’s funnel plot asymmetry test of NT-proBNP in predicting the mortality of

septic patients.
MA
E D
PT
CE
AC
 ACCEPTED MANUSCRIPT

Table 1 Characteristics of included studies

Test Survival Sensitiv
QUA
Author Count ed Cut- / ity desi Endp
Test method DAS
/year ry day off non-sur /specifi gn oint
scores
s vival city
NT-proB
NP
hospit
Roche,
Qian 401. 83.3 al
China d1 47/13 P Elecsys 2010 14
/2015 5 /80.4 morta
analyzer
lity

PT
28d Roche,
Park
Korea d2 NA 18/31 82/81 P morta Elecsys 2010 12
/2011
lity analyzer

RI
28d
Li 74.3/69
China d1 1418 53/35 P morta ELISA 13
/2011 .8

SC
lity
hospit
Roche,
Sturgess Austra al
d2 400 15/6 83/40 P Elecsys 2010 13
NU
/2010 lia morta
analyzer
lity
28d Roche,
Phua Singa 70.8
d2 6624 42/30 P morta Elecsys 2010 13
MA

/2008 pore /47.6

lity analyzer
ICU Roche,
Mokart
France d2 6624 25/26 86/77 P morta Elecsys 2010 13
/2007
lity analyzer
D

hospit
Roche,
Varpula Finlan al
E

d1 7090 187/67 45/81 P Elecsys 2010 13

/2007 d morta
analyzer
PT

lity
ICU Roche,
Roch/20 1360
France d1 22/17 73/83 P morta Elecsys 2010 13
05 0
CE

lity analyzer
Brueckm 28d
Germa 1190
ann d2 41/16 50/90 P morta ELISA 13
ny 0
/2005 lity
AC

BNP
Biosite
28d
Li 86.21/8 Triage
China d1 789 38/29 P morta 13
/2017 9.47 immunoassa
lity
y
Biosite
60d
Hu Triage
China d1 38.8 73/54 86/72.9 P morta 13
/2016 immunoassa
lity
y
28d Biosite
Yang 95.5/76
China d1 1085 39/26 P morta Triage 13
/2016 .4
lity immunoassa
 ACCEPTED MANUSCRIPT

y
Biosite
Papanik 28d
Greec Triage
olaou d1 800 20/22 65/64 P morta 13
e immunoassa
/2014 lity
y
Biosite
28d
Xing Triage
China d3 255 82/46 76.1/78 R morta 12
/2013 immunoassa
lity
y
Biosite
30d
Perman Triage

PT
USA d2 49 767/58 63/69 P morta 13
/2011 immunoassa
lity
y
hospit Biosite

RI
Sturgess Austra al Triage
d3 254 15/6 83/60 P 13
/2010 lia morta immunoassa

SC
lity y
Biosite
28d
Chen Triage
China d1 113 707/293 86/55 P morta 13
NU
/2009 immunoassa
lity
y
Biosite
28d
Zhao 681. 91.4/80 Triage
MA

China d1 63/39 P morta 13

/2009 4 .3 immunoassa
lity
y
Biosite
30d
Post/200 Germa 76.3/52 Triage
D

d5 121 55/38 p morta 13

8 ny .7 immunoassa
lity
y
E

ICU Immunoradi
Yucel Turke
PT

d1 32.1 20/20 100/95 P morta ometric 13

/2008 y
lity assay
Biosite
28d
CE

Rivers Triage
USA d1 210 214/38 79/59 P morta 13
/2007 immunoassa
lity
y
AC

28d Immunoradi
Ueda
Japan d2 650 22/12 92/80 P morta ometric 13
/2006
lity assay
Charpent 28d Immunoradi
ier France d1 190 24/10 70/67 P morta ometric 13
/2004 lity assay,
P: prospective; R: Retrospective; ICU: intensive care unit
Table 2 Subgroup analysis and meta-regression analyses of BNP
Subgroup nstudies Sensitivity P1 Specificity P2
Tested 0.86(95%CI: 0.73(95%CI:
Day one 9 0.22 0.19
day 0.79-0.92) 0.64-0.82)
 ACCEPTED MANUSCRIPT

0.79(95%CI: 0.72(95%CI:
Non-day one 5
0.68-0.91) 0.59-0.86)
Tested Immunoradiometric 0.82(95%CI: 0.70(95%CI:
11 0.05 0.02
method assay 0.76-0.89) 0.62-0.77)
Biosite 0.89(95%CI: 0.86(95%CI:
3
Triageimmuno assay 0.78-1.00) 0.73-0.98)
0.82(95%CI: 0.71(95%CI:
Endpoint 28d mortality 11 0.02 0.04

PT
0.75-0.88) 0.63-0.79)
0.90(95%CI: 0.80(95%CI:
Non-28d mortality 3 .

RI
0.81-0.99) 0.66-0.94)

SC
Table 3 Subgroup analysis and meta-regression analyses of NT-proBNP
Subgroup nstudies Sensitivity
NU P1 Specificity P2
Tested Day one 4 0.69(95%CI: 0.17 0.77(95%CI: 0.52
day 0.54-0.83) 0.65-0.88)
Non-day one 5 0.74(95%CI: 0.69(95%CI:
MA

0.61-0.87) 0.57-0.82)
Tested Elecsys2010analyzer 7 0.74(95%CI: 0.93 0.72(95%CI: 0.20
D

method 0.63-0.85) 0.61-0.82)

Non-Elecsys2010analyzer 2 0.63(95%CI: 0.78(95%CI:
E

0.41-0.84) 0.61-0.94)
PT

Endpoint 28dmortality 4 0.69(95%CI: 0.23 0.71(95%CI: 0.12

0.54-0.84) 0.58-0.83)
CE

Non-28dmortality 5 0.73(95%CI: 0.75(95%CI:

0.60-0.86 0.63-0.87)
AC
 ACCEPTED MANUSCRIPT

Highlights

 Plasma BNP and NT-proBNP were used to predict the mortality for septic patients.
 The current results of BNP and NT-proBNP for the predictive value remain
controversial.
 3417 septic patients were selected for the large meta-analysis.
 Both BNP and NT-proBNP have moderate predicts value for the mortality of septic

PT
patients.
 The tested method of BNP and NT-proBNP observation endpoint influence the results

RI
of BNP.

SC
NU
MA
E D
PT
CE
AC
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7