Background: In 2017, the regimen for patients with retreatment tuberculosis (TB) changed from an eight-month treatment regimen that consisted of Isoniazid, Rifampicin, Pyrazinamide, Ethambutol, and Streptomycin (2RHZES/1RHZE/5RHE) to a six-month treatment regimen (2RHZE/4RH). Currently, there is uncertainty regarding the effect of the six-month treatment regimen on treatment outcomes across TB Control Programs. We evaluated the effect of the six-month treatment regimen on treatment success rate (TSR) and sputum smear conversion (SSC) among patients with retreatment pulmonary bacteriologically confirmed TB (PBC-TB) in Kampala, Uganda. Methods: Between November 2021 and February 2022, we abstracted data for patients with retreatment PBC-TB who had received treatment between January 2012 and July 2021 across six large TB clinics in Kampala, Uganda. The primary outcome was TSR defined as cure or treatment completed. The secondary outcome was SSC defined as a change in sputum smear test result from positive to negative. We applied doubly-robust regression estimation for causal effect analysis, with TSR as the primary outcome and SSC at 2/3, 5, and 6/8 months as secondary outcomes. We reported average causal effects using marginal odds ratios (MOR) and 95% confidence intervals (CI). Results: We analyzed data for 427 participants with an overall mean age of 34.6±10.8 years. Over half (53.9%) of the participants were treated with a six-month treatment regimen. TSR was 49.6% overall, and higher in the six-month regimen group (52.2%) than in the eight-month regimen group (46.7%). The six-month treatment regimen significantly improved the TSR compared to the eight-month treatment regimen (MOR, 1.27; 95% CI, 1.06-1.52). However, the effect of the six-month treatment regimen on SSC at (MOR, 1.01; 95% CI, 0.91-1.12), 5 months (MOR, 0.96; 95% CI, 0.88-1.04), and 6/8 months (MOR, 0.83; 95% CI, 0.19-3.51) is similar to the eight-month treatment regimen. Conclusion: The six-month treatment regimen results in a better TSR among patients with retreatment PBC-TB compared to the eight-month treatment regimen. Therefore, patients with retreatment PBC- TB who do not have rifampicin resistance should continue to receive the six-month treatment regimen.

We abstracted data from the TB unit register between November 2021 and February 2022 for all patients with retreatment PBC-TB with known rifampicin resistance status for the period January 2012 to December 2021. We used a standardized data abstraction tool in KoBo ToolBox, a web-based and smartphone data collection tool. KoBo ToolBox was integrated with quality control measures such as skip patterns, alerts, range, and legal values to ensure accuracy in data entry. The data were collected by trained and supervised research assistants, one per study site. On each day, we downloaded the data from the server, performed data queries, and provided feedback to the research assistants to improve data accuracy.

Stata version 15 onwards.