A Cluster Randomised Trial of Community-led Distribution of HIV Self-tests in Rural Malawi (HIV Self-Testing Africa [STAR])

RESEARCH QUESTION

Can community-led delivery of HIV self-tests be used to maximise public health and social benefits and reduce costs without introducing social harms?

RESEARCH AIMS AND OBJECTIVES

The broad aim is to determine the benefits, costs and safety of community-led delivery of HIVST kits in rural Malawi, with a focus on testing and linkage to care and prevention services among defined population sub-groups: men, adolescents aged 15-19 years old, and adults aged 40 years or older.

The specific objectives are to conduct a cluster randomised controlled trial to:

• Establish the effectiveness of community-led HIVST campaigns on lifetime and recent HIV testing in predefined sub-groups.
• Investigate the impact of community-led HIVST campaigns on: (a) population-level initiation of antiretroviral therapy (ART) and uptake of voluntary medical male circumcision (VMMC), (b) knowledge of HIV prevention, (c) stigma reduction
• Explore how differences in community-led HIVST campaigns affect uptake of HIVST and linkage to care and prevention and values, attitudes and behaviours around HIVST.
• Estimate the costs and cost-effectiveness of community-led HIVST campaigns compared to the standard of care (SOC).

RESEARCH DESIGN

The main study consists of a cluster-randomised trial evaluating the effectiveness of community-led HIVST campaigns on coverage of HIV testing and linkage to follow-on services compared to the Ministry of Health (MoH) SOC, facility-based HIV testing services (HTS). We are also interested in understanding the cost-effectiveness of community-led HIVST campaigns, and broader social benefits on stigma reduction.

The unit of randomisation is the Group Village Head (GVH), who are traditionally-appointed leaders that oversee a group of villages, and their catchments areas. Outcomes are measured through (i) household surveys in evaluation villages selected for each study cluster, and (ii) clinic records of patients coming from the study clusters.

For the primary outcome, measured using the households surveys, we assume that lifetime testing rates for adolescents aged 15-19 years old in the SOC arm are 35-50%, based on the recent Demographic and Health Survey. With 16 clusters per arm and 50 adolescents per cluster, we will have at least 90% power to detect a 20% absolute increase in lifetime testing using a coefficient of variation of outcomes (k) of 0.25. If k=0.3, for a 20% increase in lifetime testing, we would have 90% and 80% power with SOC testing rates of 35-40% and 45-55%, respectively. If the absolute increase in lifetime testing is lower at 15%, then for k=0.25 we would have 80% power for SOC testing rates of 35-40%. With adolescents making up 20-25% of the adult population, this will require 250 adults per cluster.

SUB-STUDIES

We will also be conducting a series of sub-studies related to the trial, including:

• Process evaluation to monitor progress of the intervention and adherence to guidelines, and understand how differences in the context-driven intervention affect HIVST uptake and linkage to care and prevention.
• Qualitative evaluation to explore which aspects of the community-led intervention were pivotal in affecting HIV testing and linkage outcomes and values, attitudes and behaviours around HIVST.
• Economic evaluation to estimate costs, cost-effectiveness and cost-benefit per individual tested, newly diagnosed as HIV positive, started on ART and circumcised.

Inclusion criteria

• Aged ≥ 15 years
• Residence in GVH catchment areas included in the intervention arm

Exclusion criteria

• Aged < 15 years old
• Not resident in GVH catchment areas included in the intervention arm