Central Nervous System Tuberculosis: Risk Factors for Mortality in a Propensity Score–Matched Case-Control Study

Abstract

Background Despite advancements in tuberculosis (TB) control and treatment in the United States (US), patients with central nervous system TB (CNS-TB) continue to experience significantly higher mortality rates than those without CNS-TB. This raises concerns regarding clinical management and the need for a deeper understanding of the risk factors contributing to these deaths. This study aimed to determine the predictors of mortality in patients with CNS-TB. Methods We conducted a retrospective 1:2 propensity-score-matched case-control study. Cases were TB patients diagnosed with TB of the meninges, brain, spinal cord, or peripheral nerves, as documented in the Florida Department of Health (FDOH) TB registry, between 2009 and 2021. Controls were TB patients without CNS-TB, also reported in the FDOH TB registry during the same timeframe. We employed conditional logistic regression models to investigate the factors contributing to mortality in cases compared with controls. Results We analyzed data from 116 cases and 232 matched controls. Patients with CNS-TB had a 5.69-fold higher risk of death than those without CNS-TB (aOR = 5.69, 95% CI: 2.91, 11.6). Increased risk of death was associated with HIV co-infection (aOR = 1.93, 95% CI: 0.82, 4.37) and diabetes (aOR = 3.13, 95% CI: 1.28, 7.47). Miliary TB and non-HIV immunosuppression were significantly associated with being a case, while cavitary TB was less likely to be associated with being a case. Conclusions Clinical management should prioritize screening and close monitoring of patients with HIV co-infection and diabetes to improve patient outcomes.

Numerous Asymptomatic Brain Tuberculomas Complicated by Fatal Tuberculous Meningitis

“…This problem with unfavorable outcomes persists in locations with higher resources, where mortality and neurologic sequelae rates can still reach 50%. One reason is the laboriousness and limited sensitivity of available diagnostic tools, which may result in diagnosis when CNS involvement is too advanced for effective treatment [12].…”