Mortality hazards after treatment completion of pulmonary tuberculosis in a tertiary hospital in Uganda

Abstract

Objectives

Treated pulmonary tuberculosis (PTB) is associated with long-term complications that contribute to substantial morbidity and mortality. We estimated the incidence and predictors of post-PTB mortality and evaluated whether chronic pulmonary aspergillosis (CPA) independently increases the risk of death.

Methods

Between July 1, 2020 and June 30, 2021, we enrolled 162 individuals with treated drug-susceptible PTB who had persistent respiratory symptoms and screened them for CPA using a symptom checklist, chest x-ray, Aspergillus immunoglobulin G-immunoglobulin M point-of-care test, and sputum culture. Between November and December 2024, we followed up all participants via phone calls to determine their vital status. On chest X-ray, PTB was classified as minimal, moderate, and far advanced disease based on involvement of one, two, or more zones, respectively, and coupled to unilateral or bilateral lung disease. Cox proportional hazards regression was used to identify independent predictors of mortality.

Results

Thirty-seven (22.8%) participants were lost to follow-up. The median follow-up duration was 3.8 years (interquartile range 3.6-3.9). Of the 125 participants with vital status, their mean age was 33.5 years (±11.7). At baseline, 46 (36.8%) had far advanced PTB, 64 (51.2%) had pulmonary fibrosis, 15 (12.0%) had a history of previous TB, and 34 (27.2%) were living with HIV. Coinfection with PTB and CPA was identified in 31 participants at baseline (24.8%). The median St. George’s respiratory questionnaire score was 50.9 (interquartile range 40.9-63.3), and 32.0% (n = 40) had scores above 60, indicating poor health-related quality of life. Overall mortality was 8.8% (95% confidence interval [CI] 4.4-15.2%), with a mortality rate of 24.3 deaths per 1000 person-years of follow-up. Mortality rates were comparable among participants with and without CPA-PTB coinfection. Independent predictors of mortality included a St. George’s respiratory questionnaire score >60 (adjusted hazard ratio = 2.01; 95% CI 1.49-2.72; P <0.001) and HIV infection (adjusted hazard ratio = 3.04; 95% CI 1.46-6.34; P = 0.029).

Conclusions

Post-PTB mortality remains high, with poor health-related quality of life and HIV co-infection emerging as significant independent predictors of death. Integrating long-term follow-up, respiratory rehabilitation, and fungal diagnostics into post-TB care pathways is essential to improve outcomes and reduce preventable mortality.