Sustaining progress in HIV care in Africa with a transition to enabled self-care: a modelling study

Background: The funding crisis for HIV services in east, central, southern, and west Africa means that innovative strategies for continuing prevention and care for HIV are needed. One such strategy is enabled self-care: the provision of free HIV self-tests, pre-exposure prophylaxis (PrEP), tenofovir–lamivudine–dolutegravir for post-exposure prophylaxis (PEP), and antiretroviral therapy (ART) provided in local pharmacies and by community-based health workers. We modelled the introduction of a policy of transition to enabled self-care compared with continuation of current service provision. 

Methods: We used an established individual-based model (HIV Synthesis) to simulate HIV epidemics in 1000 setting scenarios representing diversity in setting characteristics across east, central, southern, and west Africa, and uncertainty in parameter values. For each setting scenario, we simulated 100 000 people aged ≥15 years between 1989 and 2076. The model incorporates ART access informed by studies of barriers to accessing clinics. Our main outcomes included the number of adults testing for HIV in a 3-month period, PrEP and PEP use, coverage of people on ART, HIV-related deaths, HIV incidence, perinatal transmission, disability-adjusted life-years (DALYs), and costs. We chose a 50-year time horizon to capture the full effects of the policies. For each setting scenario, we performed three replicate runs for each of the two policies and for each policy took the mean across the three runs. 

Findings: Our model estimated that transitioning to enabled self-care would lead to increases in testing from a median of 5·0% (90% range 2·3–11·5) of adults testing per 3 months to 12·7% (6·6–21·1), an increase in PrEP and PEP use (+0·5% of adults [0·0–2·2]), and an increase in ART coverage for people with diagnosed HIV (+1% [0–3]) with enhanced benefits in those previously without access to PrEP and testing services. Enabled self-care was predicted to result in a median 18% (90% range 4–33) fewer HIV deaths, and a relative rate of 0·73 (0·52–0·94) in terms of HIV incidence per 100 person-years over 10 years. Overall, over a 50-year time horizon, enabled self-care was predicted to cut discounted programme costs by US$6·0 million (5·6–6·5; a 7% decrease), avert 34 000 (32 200 to 36 000) DALYS, and be cost-effective (at a cost-effectiveness threshold of $300 per DALY averted) in 95% of setting scenarios. The mean undiscounted annual cost per adult in the population per year across setting scenarios was $12 (median $11 [90% range 6–22]). The higher the HIV prevalence was in a setting scenario, the more cost-effective the policy was found to be (odds ratio 1·41 [95% CI 1·25–1·60] per 1% higher HIV prevalence). 

Interpretation: Introduction of community access to self-tests and antiretroviral drugs through a transition to enabled self-care is very likely to be cost-effective in most settings in east, central, southern, and west Africa; enable incidence declines to be sustained at reduced cost; and increase equity of access to HIV services. Policy makers who wish to consider such a policy will need to evaluate its feasibility in their own country settings. 

Funding: UCL and Gates Foundation.