A New Scoring System for the Differential Diagnosis between Tuberculous Meningitis and Viral Meningitis

Background: Tuberculous meningitis (TBM) is associated with high mortality and morbidity despite administering anti-tuberculous chemotherapy to the patients. Differential diagnosis between TBM and viral meningitis (VM) is difficult in some clinical situations. Methods: We reviewed and analyzed records of adult patients who were admitted and diagnosed with TBM or VM at a tertiary hospital in Korea, between January 2006 and December 2015. Diagnostic criteria for TBM were categorized into three groups: definite, probable, and possible TBM. The VM group included patients with no evidence of other meningitis who achieved complete recovery with only conservative treatments. Clinical, laboratory and radiological findings, as well as outcomes, were compared between the TBM and VM groups. Results: Ninety-eight patients were enrolled. Among the study patients, 47 had TBM and 51 had VM. Based on univariate analysis and multivariate logistic regression, sodium 70 (U/L) in cerebrospinal fluid (CSF), protein > 160 (mg/dL) in CSF, voiding difficulty, and symptoms of cranial nerve palsy were significant predictive factors for TBM in the final model. We constructed a weighted scoring system with predictive factors from multiple regression analyses. Receiver operating characteristic curve analyses and decision tree analyses were plotted to reveal an optimum cutoff point as 4 with this scoring system (range: 0-13). Conclusion: For differential diagnosis between TBM and VM, we created a new weighted scoring system. This scoring system and decision tree analysis are simple and easy to apply in clinical practice to differentiate TBM from VM.