Can HIV care changes affect TB prevention?

People living with HIV face a higher risk of developing tuberculosis, so programs often offer tuberculosis preventive therapy (TPT) alongside antiretroviral therapy (ART). As health systems in sub-Saharan Africa scale up differentiated service delivery (DSD) models — which reduce clinic visits and change how ART is delivered — researchers have asked whether moving patients into these new models affects other services such as TPT. Ann Johnson and colleagues examined this question using data from Opt4TPT, a longitudinal cohort study conducted in South Africa and Zimbabwe. Using the RE-AIM framework to guide their work, the team set out to compare uptake and completion of TPT between people receiving ART through DSD models and those remaining in conventional care. The study tracked when people started TPT and whether they completed the therapy, with the goal of understanding whether DSD programs can support not only ART delivery but also essential prevention measures against tuberculosis.

The study included 1,193 participants, of whom 276 received ART through a DSD model and 917 through conventional care. Overall, 1,035 (87%) initiated TPT: 242 (88%) in DSD models and 793 (86%) in conventional models. Multivariate logistic regression found no significant association between receiving ART in a DSD model and TPT initiation (OR 1.11, 95% CI 0.74-1.67, p = 0.61). However, time to initiation differed: DSD participants had a mean time to TPT start of 6.5 days versus 2.7 days in conventional care (p = 0.01), a difference assessed with a Cox proportional hazards model. Electronic monitoring using MERM box data was available for 731 participants; 356 (49%) completed TPT. In bivariate analysis, those in DSD models had higher odds of completing TPT (OR 1.53, 95% CI 1.06-2.21, p = 0.024), but this association disappeared in multivariate analysis after adjusting for demographic and clinical factors (OR 0.89, 95% CI 0.58-1.36, p = 0.58).

Taken together, the results show that offering TPT within DSD models is feasible: uptake was high in both DSD and conventional care, with the majority of participants starting preventive therapy. At the same time, the overall completion rate was low, with less than half of monitored participants finishing TPT, and the apparent advantage for DSD in simple comparisons did not hold after accounting for other factors. The longer average time to initiation in DSD models suggests operational differences that programs should understand and address. For health systems scaling DSD, these findings imply that adding TPT to DSD is possible but that extra attention is needed to support people through the full course. The study highlights a clear need to focus on strategies to improve TPT completion in both DSD and conventional models to better protect people living with HIV from tuberculosis.