Cerebrospinal Procalcitonin for Diagnosing Meningitis in Infants Less Than 90 Days: A Systematic Review and Meta-analysis.

Context: Early identification and treatment of meningitis in early infancy is crucial for improved outcomes. However, the commonly used cerebrospinal fluid (CSF) markers for diagnosing meningitis have poor diagnostic accuracy.

Objective: To assess the diagnostic utility of CSF procalcitonin for diagnosing meningitis in infants with postnatal age less than 90 days.

Evidence acquisition: We searched databases, including MEDLINE, Embase, and Scopus, from inception to 15 March 2025. We included observational studies that used CSF procalcitonin as a marker for diagnosing meningitis in infants with a postnatal age of less than 90 days. The risk of bias was assessed using the QUADAS-2, and the certainty of evidence was ascertained using GRADE approach. The bivariate random effect model and hierarchical summary receiver operating characteristic meta-analysis were used to pool data with multiple thresholds.

Results: Out of 815 retrieved records, seven studies involving 963 infants were included in the meta-analysis. We obtained diagnostic accuracy measures at commonly used threshold cut-offs of 0.20 ng/mL and 0.33 ng/mL. At a CSF procalcitonin threshold cut-off of 0.20 ng/mL, the pooled sensitivity (95%CI) was 77.1% (49.2%, 92.1%), and the pooled specificity (95%CI) was 68.9% (49.2%, 83.5%). At a cut-off of 0.33 ng/mL, sensitivity was 75.8% (50.1%, 90.7%) and specificity was 75.1% (58.5%, 86.6%). The certainty of evidence remained very low for both sensitivity and specificity.

Conclusion: CSF procalcitonin has moderate sensitivity and specificity for diagnosing meningitis in infants with postnatal age < 90 days. However, wide confidence intervals and very low certainty of evidence limit the reliability of these findings.