Cerebrospinal fluid procalcitonin and neutrophil percentage: a combined biomarker for differentiating bacterial from tuberculous meningitis in antibiotic-pretreated patients.

Background: Distinguishing between bacterial meningitis (BM) and tuberculous meningitis (TBM) is a persistent clinical dilemma, further complicated by widespread empirical antibiotic use. This study evaluated the diagnostic and prognostic utility of cerebrospinal fluid procalcitonin (CSF-PCT) and other laboratory markers, analyzing all enrolled patients irrespective of prior antibiotic exposure. Methods: A prospective cohort analysis was conducted from September 2014 to September 2025, consecutively enrolling 125 BM patients and 56 TBM patients. Patients were included regardless of their pre - admission antibiotic treatment history. Demographic and laboratory parameters upon admission were analyzed. Receiver operating characteristic and multivariate logistic regression analyses were employed to identify biomarkers for differentiating BM from TBM and predictors of unfavorable outcomes (Glasgow Coma Scale score ≤8 at discharge) in BM. Results: The majority (89.5%) of patients had prior antibiotics. CSF-PCT and CSF neutrophil percentage demonstrated comparable efficacy in distinguishing BM from TBM, with area under the curve (AUC) values of 0.772 and 0.786, respectively. A combined model integrating both biomarkers achieved superior diagnostic performance (AUC = 0.849), significantly outperforming either marker alone (both Conclusion: The combination of CSF-PCT and CSF neutrophil percentage serves as a reliable biomarker for differentiating BM from TBM. CSF-PCT stands as an independent risk factor for unfavorable short-term outcomes in BM. These findings confirm the clinical utility of these biomarkers in early diagnosis and risk stratification, which remains effective in the routine clinical setting where patients commonly present with a history of antibiotic use.