When determining the effectiveness of a new treatment, randomized trials are not always possible or desirable. The stability-controlled quasi-experiment (SCQE) (Hazlett, 2019) is an observational approach that replaces the usual “no-unobserved confounding” assumption with one on the change in non-treatment outcome between successive cohorts, or the “baseline trend.” We extend this method to allow variance estimation and inference, and apply it for the first time by examining whether isoniazid preventive therapy (IPT) reduced tuberculosis (TB) incidence among 26,715 HIV patients in Tanzania. After IPT became available in the clinics we studied, a non-random 25% of patients received it. Within a year, fewer than 1% of patients on IPT developed TB, compared to 16% of the untreated. Regression adjustment using available covari-ates produces an equally large and highly significant estimate of -15 percentage point (pp) [95%CI: -16.6, -13.7]. While those estimates may generate confidence in IPT’s effectiveness, they cannot eliminate confounding. By contrast, SCQE reveals that the average treatment effect on the treated must be small and indistinguishable from zero, if we assume the baseline trend was flat over the study period. Rather, to argue that IPT was beneficial requires claiming that the (non-treatment) incidence rate rose by at least 0.5 pp per year. This is plausible, but far from certain. The SCQE approach has broad applicability and will sometimes lead to definitive claims of effectiveness. In this case, it usefully aids in protecting against over-confidence in claims that IPT was effective.

PURPOSE OF THE REVIEW: In 2015, antiretroviral therapy (ART) was recommended for all people living with HIV (PLHIV) regardless of CD4 count (Treat All). To better understand how to improve linkage to care under these new guidelines, we conducted a systematic review of studies evaluating linkage interventions in Sub-Saharan Africa under Treat All. RECENT FINDINGS: We identified 14 eligible articles and qualitatively analyzed the effectiveness of the interventions. Increases in linkage were reported by supply-side and counseling interventions. Mobile testing and economic incentives did not increase linkage. Given the lag time between adoption and implementation, only two of the studies were conducted in a Treat All setting. None of the interventions specifically focused on re-linking PLHIV who had disengaged from care. Future studies must design interventions that target not only newly diagnosed or treatment naïve PLHIV, but should explicitly focus on PLHIV who have disengaged from care.

Introduction

Adverse sexual and reproductive health (SRH) outcomes, such as unplanned pregnancies and HIV infection, disproportionately affect adolescent girls and young women (AGYW; aged 15-24 years) in east Africa. Increasing uptake of preventive SRH services via innovative, youth-centred interventions is imperative to addressing disparities in SRH outcomes.

Methods and analysis

From 2018 to 2019, we used human-centred design to co-develop a theoretically driven HIV and pregnancy prevention intervention for AGYW at private drug shops called Accredited Drug Dispensing Outlets (ADDOs) in Tanzania. The result, Malkia Klabu (Queen Club), was a customer loyalty programme designed to strengthen ADDOs' role as SRH providers while encouraging uptake of critical SRH prevention products among AGYW. Malkia Klabu members had access to free contraceptives and oral HIV self-test (HIVST) kits and earned punches on a loyalty card for other shop purchases; punches were redeemable for small prizes. Our pilot among 40 shops showed that intervention ADDOs had higher AGYW patronage and distributed more HIVST kits and contraceptives to AGYW relative to business-as-usual (ie, client purchasing) comparison shops. We will conduct a cluster-randomised controlled trial (c-RCT) among 120-140 ADDOs in 40 health catchment areas in Shinyanga and Mwanza Regions (Lake Zone), Tanzania. ADDO shop recruitment includes a 1-month run-in with a tablet-based electronic inventory management system for tracking shop transactions, followed by enrolment, randomisation and a 24-month trial period. Our c-RCT evaluating the human-centred design-derived intervention will assess population impact on the primary outcomes of HIV diagnoses and antenatal care registrations, measured with routine health facility data. We will also assess secondary outcomes focusing on mechanisms of action, evaluate programme exposure and AGYW behaviour change in interviews with AGYW, and assess shop-level implementation strategies and fidelity.

Ethics and dissemination

Ethical approval was granted from both the University of California, San Francisco and the Tanzanian National Institute for Medical Research. Study progress and final outcomes will be posted annually to the National Clinical Trials website; study dissemination will occur at conferences, peer-reviewed manuscripts and local convenings of stakeholders.

Trial registration number

NCT05357144.

OBJECTIVES: This study aimed to develop a machine learning (ML) model to predict disengagement from HIV care, high viral load or death among people living with HIV (PLHIV) with the goal of enabling proactive support interventions in Tanzania. The algorithm addressed common challenges when applying ML to electronic medical record (EMR) data: (1) imbalanced outcome distribution; (2) heterogeneity across multisite EMR data and (3) evolving virological suppression thresholds. DESIGN: Observational study using a national EMR database. SETTING: Conducted in two regions in Tanzania, using data from the National HIV Care database. PARTICIPANTS: The study included over 6 million HIV care visit records from 295 961 PLHIV in two regions in Tanzanias National HIV Care database from January 2015 to May 2023. RESULTS: Our ML model effectively identified PLHIV at increased risk of adverse outcomes. Key predictors included past disengagement from care, antiretroviral therapy (ART) status (which tracks a patients engagement with ART across visits), age and time on ART. The downsampling approach we implemented effectively managed imbalanced data to reduce prediction bias. Site-specific algorithms performed better compared with a universal approach, highlighting the importance of tailoring ML models to local contexts. A sensitivity analysis confirmed the models robustness to changes in viral load suppression thresholds. CONCLUSIONS: ML models leveraging large-scale databases of patient data offer significant potential to identify PLHIV for interventions to enhance engagement in HIV care in resource-limited settings. Tailoring algorithms to local contexts and flexibility towards evolving clinical guidelines are essential for maximising their impact.