Development and validation of a new scoring system for the early diagnosis of tuberculous meningitis in adults

We developed and validated a new diagnostic scoring system by simultaneously comparing 28 factors (including clinical, laboratory, and imaging) of HIV uninfected adult tuberculous meningitis (TBM) with viral meningitis (VM), bacterial meningitis (BM), and cryptococcal meningitis (CM). Predictors of TBM diagnosis obtained by logistic regression. A total of 382 patients with intracranial infection participated, and eight factors were independently associated with TBM diagnosis: symptom duration, evidence of extracranial tuberculosis, cerebrospinal fluid (CSF) leukocyte, CSF neutrophil, CSF protein, low serum sodium, meningeal enhancement, and tuberculomas. Factors are assigned according to weight, a score of ≥ 5 was suggestive of TBM with a sensitivity of 85.8% and a specificity of 87.7%, and the area under the receiver operating characteristic curve was 0.923. When applied to a prospective validation cohort, this scoring model showed robust performance. Our study suggests that the application of this score can help diagnose TBM more efficiently.