RESUMO
Tuberculous meningitis (TBM) is a medical emergency. Early and accurate diagnosis is crucial for preventing morbidity and mortality. Our aim was to define the predictors of tuberculous rather than other causes of acute aseptic meningitis syndrome (AAMS). A retrospective study of 225 patients with AAMS admitted in a tertiary care infectious diseases center between 2001 and 2004 was performed. In order to assess the value of clinical and biological variables for the diagnosis of TBM, we compared the variables between two groups of patients, one with microbiologically documented TBM and one with certain non-TBM acute lymphocytic meningitis. The assessed variables had good specificities, but relatively low sensitivities, ruling in the diagnosis of TBM rather than ruling it out. The most reliable predictors in multivariate analysis were: duration of symptoms before admission > 7 days, altered clinical stage, CSF/blood glucose ratio < 0.5 and CSF protein concentration > 200mg/ml. Based on the multivariate model which retained four variables, the probability of TBM in a certain patient could have been increased from 17% to 99.9% when all variables were present, or decreased to 0% when all variables were absent. This study suggests that simple clinical and laboratory features are useful in diagnosing TBM in a high-prevalence region.
