Diagnostic Prediction Model for Tuberculous Meningitis: An Individual Participant Data Meta-Analysis.

No accurate and rapid diagnostic test exists for tuberculous meningitis (TBM), leading to delayed diagnosis. We leveraged data from multiple studies to improve the predictive performance of diagnostic models across different populations, settings, and subgroups to develop a new predictive tool for TBM diagnosis. We conducted a systematic review to analyze eligible datasets with individual-level participant data (IPD). We imputed missing data and explored three approaches: stepwise logistic regression, classification and regression tree (CART), and random forest regression. We evaluated performance using calibration plots and C-statistics via internal-external cross-validation. We included 3,761 individual participants from 14 studies and nine countries. A total of 1,240 (33%) participants had "definite" (30%) or "probable" (3%) TBM by case definition. Important predictive variables included cerebrospinal fluid (CSF) glucose, blood glucose, CSF white cell count, CSF differential, cryptococcal antigen, HIV status, and fever presence. Internal validation showed that performance varied considerably between IPD datasets with C-statistic values between 0.60 and 0.89. In external validation, CART performed the worst (C = 0.82), and logistic regression and random forest had the same accuracy (C = 0.91). We developed a mobile app for TBM clinical prediction that accounted for heterogeneity and improved diagnostic performance (https://tbmcalc.github.io/tbmcalc). Further external validation is needed.

Diagnostic and Prognostic Value of Cerebrospinal Fluid Lactate and Glucose in HIV-Associated Tuberculosis Meningitis.

The role of cerebrospinal fluid (CSF) lactate in tuberculosis meningitis (TBM) diagnosis and prognosis is unclear. The aim of this study was to evaluate the performance of CSF lactate alone and in combination with CSF glucose in predicting a diagnosis of TBM and 14-day survival. HIV-positive Ugandan adults were investigated for suspected meningitis. The baseline CSF tests included smear microscopy; Gram stain; cell count; protein; and point-of-care glucose, lactate, and cryptococcal antigen (CrAg) assays. Where CrAg was negative or there was suspicion of TBM, a CSF Xpert MTB/RIF Ultra (Xpert Ultra) test was performed. We recorded baseline demographic and clinical data and 2-week outcomes. Of 667 patients, 25% (n = 166) had TBM, and of these, 49 had definite, 47 probable, and 70 possible TBM. CSF lactate was higher in patients with definite TBM (8.0 mmol/L; interquartile ratio [IQR], 6.1 to 9.8 mmol/L) than in those with probable TBM (3.4 [IQR, 2.5 to 7.0] mmol/L), possible TBM (2.6 [IQR 2.1 to 3.8] mmol/L), and non-TBM disease (3.5 [IQR 2.5 to 5.0] mmol/L). A 2-fold increase in CSF lactate was associated with 8-fold increased odds of definite TBM (odds ratio, 8.3; 95% confidence interval [CI], 3.6 to 19.1; P < 0.01) and 2-fold increased odds of definite/probable TBM (odds ratio, 2.3; 95% CI, 1.4 to 3.7; P < 0.001). At a cut point of >5.5 mmol/L, CSF lactate could be used to diagnose definite TBM with a sensitivity of 87.7%, specificity of 80.7%, and a negative predictive value of 98.8%. CSF lactate was not predictive of 2-week mortality. IMPORTANCE Tuberculosis meningitis (TBM) is the most severe form of tuberculosis, and its fatality is largely due to delays in diagnosis. The role of CSF lactate has not been evaluated in patients with HIV presenting with signs and symptoms of meningitis. In this study, using a point-of-care handheld lactate machine in patients with HIV-associated meningitis, we showed that high baseline CSF lactate (>5.5 mmol) may be used to rapidly identify patients with TBM and shorten the time to initiate treatment with a similar performance to the Xpert Ultra assay for definite TBM. Elevated CSF lactate levels, however, were not associated with increased 2-week mortality in patients with HIV-associated TBM. Due to moderate specificity, other etiologies of meningitis should be investigated.