Trends in prevalent TB among persons enrolling for HIV care before and after 'Test and Treat' across East-Africa

Abstract

Background: In 2015, WHO recommended the global adoption of the 'Test and Treat' strategy (TTS) for all persons living with HIV (PLHIV). While TTS has improved viral suppression and reduced mortality, its impact on TB in PLHIV remains unclear.

Methods: We assessed TB prevalence trends 48 months before and after TTS among PLHIV aged ≥18 years enrolling at HIV primary care sites affiliated with the East Africa International Epidemiology Databases to Evaluate AIDS (EA-IeDEA) consortium. We defined prevalent TB as bacteriologically confirmed or empirically treated TB within 60 days of enrolment. We estimated monthly TB prevalence trends using Poisson (change point) model.

Results: Among 125,647 PLHIV, 37% were male. The prevalence of TB was 8.9% (95% CI: 8.7-9.1) before and 6.2% (95% CI: 5.9-6.4) after TTS-adoption. Adjusted analysis showed significant downward trend in TB prevalence before TTS (adjusted Prevalence Rate Ratio, aPRR=0.989, p<0.001), which plateaued during TTS (aPRR=0.999, p=0.131). TB was more frequently present among males (aPRR: 2.09, p<0.001) and adults ≥25 years across both periods.

Conclusion: This study highlights a plateau in TB prevalence decline during TTS and persistent disparities in TB by sex and age, underscoring the need for targeted interventions.

Keywords: ART; Kenya; PLHIV; Tanzania; Uganda; prevalence; tuberculosis.

Figure 1.

Study Flow Chart showing PLHIV enrolled into care months before and after Test and Treat strategy (TTS) implementation.

Figure 2.

Change point model showing EA trends in TB prevalence before and after Test and Treat strategy (TTS) implementation. A smooth linear trend (blue dashed line) was added to assess model fit. The orange dashed line shows what TB prevalence would have been, had the trend of TB prevalence over time, observed prior to TTS initiation, continued after inception of the policy. The trends are shown in solid black lines in the linear scale. A smooth trend was also fit. It largely follows the linear trend posited by the model, supporting the linear structural assumption imposed by the CP model as well as the lack of any visible continued reduction in TB prevalence post TTS.

Figure 3.

Trends in TB prevalence by country before and after Test and Treat strategy (TTS) implementation. The figure shows trends of monthly TB prevalence among PLHIV enrolled in IeDEA cohort. The dotted line indicates a point when TTS was rolled out. The prevalence of TB was on a declining trend across all three countries before TTS. Following TTS, the TB prevalence trend in Tanzania continued to decline, but plateaued in Kenya and went up in Uganda.