Diagnostic performances of procalcitonin and C-reactive protein for sepsis: a systematic review and meta-analysis

BACKGROUND: The Sepsis-3 2016 definition defined sepsis as a life-threatening organ dysfunction caused by a dysregulated host response to infection. Procalcitonin (PCT) and C-reactive protein (CRP) have been widely studied for the detection of sepsis according to the former definitions. This study aimed to evaluate the diagnostic performances of PCT and CRP for sepsis, according to the Sepsis-2 and Sepsis-3 definitions. METHODS: PubMed, Embase, and the Cochrane Library were searched. Original articles that reported both diagnostic performances of PCT and CRP for sepsis were included. The pooled sensitivity, specificity, diagnostic odds ratio, likelihood ratio, and the area under the summary receiver operating characteristic curve (AUC) were calculated using the multiple thresholds model. RESULTS: Forty-four studies with 10 755 patients between 1997 and 2024 were included. PCT exhibited a higher pooled AUC of 0.74 [95% confidence interval (CI), 0.62-0.84] compared with CRP, which had an AUC of 0.67 (95% CI, 0.56-0.77). Using sensitivity weighting of 50%, the optimal PCT and CRP cutoffs were 0.54 ng/ml (sensitivity: 0.70; specificity: 0.67) and 48 mg/L (sensitivity: 0.72; specificity: 0.55), respectively. The pooled AUC of PCT did not significantly differ between the Sepsis-2 and Sepsis-3 criteria. Sensitivity analyses showed overall performance was higher using the traditional bivariate model than the multiple thresholds model. CONCLUSIONS: Although PCT seems to slightly outperform CRP for the diagnosis of sepsis, its discriminatory power remains limited, highlighting the need for additional tools to improve sepsis diagnosis.