Time to pulmonary tuberculosis relapse and its association with chest radiograph in Kawempe division, Kampala

Abstract

Introduction: In Uganda, 5.19% of new tuberculosis cases are relapses with a prevalence and incidence of 0.159% and 0.161% respectively. Few studies have assessed factors that act as biomarkers for relapse at the initial episodes of PTB so that prolonged therapy can be apportioned to individuals with these factors. If these factors are not identified, there is a possibility of poor adherence to treatment during the second episode hence increased likelihood of MDR development. Objectives: To determine time to relapse and assess its association with chest radiographic pattern among patients initially diagnosed with TB and who completed treatment and were bacteriologically confirmed cured at the end of treatment. Methods: A retrospective cohort study design was used to assess time to relapse among index cases in the Kawempe Community Health (KCH) prospective cohort study who were followed up for 18 months after successful completion of standard TB chemotherapy. Results: The incidence rate of relapse was 9.4 per 100 person years [95% CI 6.33-11.18] among 669 sputum smear or culture confirmed patients cured of their initial PTB disease. The median time to relapse was 363 days. There was no difference in the median time to relapse of patients with normal/minimal radiological extent of disease and those with advanced radiological extent of disease. Patients aged 28 years and above were more like to relapse, HR 2.14 [95% CI 1.01 4.17]. Conclusion and Recommendations: Although chest radiographic patterns are not good predictors of TB relapse, physicians should consider prolonging chemotherapy for older individuals (more than 27 years of age) with persistent radiographic TB fibrosis at the end of short course (6 months) chemotherapy.