OBJECTIVE: To examine the prevalence and magnitude of chronic lung disease (CLD) and its association with empiric anti-tuberculosis treatment (due to lack of bacteriologic confirmation) among recurrent tuberculosis (TB) survivors in a human immunodeficiency virus (HIV) prevalent setting.
METHODS: Prospective cohort study of retreatment TB survivors in Harare, Zimbabwe. At median follow-up of 2 years post-treatment initiation, we characterized mortality, respiratory impairment, and mental health.
RESULTS: Among 175 retreatment TB survivors, 65% of whom were HIV-positive and 21% had been empirically treated, multiparameter CLD was noted at follow-up among 14% of patients (95%CI 9.0-19.7), with a six-fold increase in age-adjusted death in the first year following treatment completion. Empirically treated TB (relative risk [RR] 3.4, 95%CI 1.4-8.3) was associated with CLD, as was the number of previous anti-tuberculosis treatment courses in dose-dependent fashion (three vs. one, RR 6.2, 95%CI 1.7-22.1). Among retreatment TB survivors, 33% (95%CI 26.0-40.1) had persistent respiratory symptoms (Chronic Obstructive Pulmonary Disease Assessment Test score >/=10); 26% (95%CI 19.8-33.0) significant deficits in exercise capacity (median incremental shuttle walk test distance, 550 m; Q1-Q(3) 440-730 m); 83% (95%CI 75.7-89.7) residual radiographic abnormalities on chest X-ray; 12% (95%CI 6.6-16.1%) moderate-to-severe obstruction on spirometry; and 13% (95%CI 7.6-17.5%) major depression.
CONCLUSIONS: Despite successful treatment, retreatment TB survivors retain a substantial risk of morbidity and mortality.