Natriuretic Peptides to Predict Short-Term Mortality in Patients With Sepsis: A Systematic Review and Meta-analysis

Area of Science:

• Biomarkers in Critical Care
• Sepsis Pathophysiology
• Clinical Diagnostics

Background:

• Conflicting data exists on optimal B-type natriuretic peptide (BNP) and N-terminal pro-B-type natriuretic peptide (NT-proBNP) cutoffs for predicting short-term mortality in sepsis.
• Accurate prediction of sepsis-related mortality is crucial for timely clinical intervention.

Purpose of the Study:

• To identify optimal BNP and NT-proBNP cutoffs for predicting short-term mortality in adult patients with sepsis.
• To synthesize evidence from existing studies to establish reliable predictive values.

Main Methods:

• A comprehensive literature search was conducted across multiple databases (MEDLINE, EMBASE, Cochrane, Scopus) for relevant English-language studies published between 2000 and 2017.
• Thirty-five observational studies involving 3508 patients with sepsis, severe sepsis, or septic shock were included.
• Receiver operating characteristic (ROC) curves were estimated using a parametric regression model based on average values in survivors and nonsurvivors.

Main Results:

• A BNP cutoff of 622 pg/mL demonstrated the greatest discrimination for mortality (AUC, 0.766).
• An NT-proBNP cutoff of 4000 pg/mL showed the highest discrimination for mortality (AUC, 0.787).
• Higher predictive accuracy was observed for BNP/NT-proBNP cutoffs obtained within 24 hours of admission, in severe sepsis/septic shock, and in studies from the US and Europe.

Conclusions:

• BNP cutoff of 622 pg/mL and NT-proBNP cutoff of 4000 pg/mL are proposed as optimal predictors of short-term mortality in sepsis.
• These findings require further validation due to heterogeneity in patient populations and inconsistent adjustment for renal function in the included studies.