Missed Opportunities in the Diagnosis of Tuberculosis Meningitis

Abstract

Background Tuberculosis meningitis (TBM) has high mortality and morbidity. Diagnostic delays can impact TBM outcomes. We aimed to estimate the number of potentially missed opportunities (MO) to diagnose TBM, and determine its impact on 90-day mortality. Methods Retrospective cohort of adult patients with a central nervous system (CNS) TB ICD-9/10 diagnosis code (013*, A17*) identified in the Healthcare Cost and Utilization Project, State Inpatient and State Emergency Department (ED) Databases from eight states. MO was defined as composite of ICD-9/10 diagnosis/procedure codes that included CNS signs/symptoms, systemic illness, or non-CNS TB diagnosis during a hospital/ED visit 180 days prior to the index TBM admission. Demographics, comorbidities, admission characteristics, mortality, and admission costs were compared between those with and without a MO, and 90-day in-hospital mortality, using univariate and multivariable (MV) analyses. Results Of 893 patients with TBM median age at diagnosis was 50 years (IQR 37, 64), 61.3% were male, and 35.2% had Medicaid as primary payer. Overall, 407 (45.6%) had a prior hospital or ED visit with a MO code. In-hospital 90-day mortality was not different between those with and without a MO, regardless of the MO coded during an ED visit (13.7% vs 15.2%, p=0.73) or a hospitalization (28.2% vs 30.9%, p=0.74). Independent risk of 90-day in-hospital mortality was associated with older age, hyponatremia (RR 1.62, 95%CI 1.1–2.4, p=0.01), septicemia (1.6, 1.03–2.45, p=0.03), and mechanical ventilation (3.4, 2.25–5.3, p<0.001) during the index admission. Conclusion Almost half the patients coded for TBM had a hospital or ED visit in the previous 6 months meeting the MO definition. We found no association between having a MO for TBM and 90-day in-hospital mortality.