Economic evaluation of models of prevention of mother-to-child transmission of HIV intervention for large scale implementation

Background: Huge successes have been seen in the prevention of mother-to-child transmission of HIV (PMTCT) towards its elimination. Now amidst a landscape of universal antiretroviral therapy (ART), focus has been placed on different models of care to support and retain mother-infant pairs in the vulnerable postpartum phase. Methods The aim was to establish economic evidence for scaling-up approaches and models of care for PMTCT particularly during the postpartum period in Southern Africa. The economic data were collected during three studies, Safe Generations (Eswatini), MCH-ART and PACER (South Africa), using mixed bottom-up and top-down methodology. Outcomes of these studies were used to estimate the cost-effectiveness using an incremental cost effectiveness ratio (ICER, calculated by the difference in cost divided by the difference in effects) of lifelong ART in comparison to Option A (the standard of care at the time) in Eswatini; and to estimate the annual costs, costeffectiveness and budget impact of three models of care (Model I: Routine Care - mothers in general ART and infants in well-baby clinics; Model II: Integrated Care - mothers-infant pairs in integrated care in midwife obstetric unit; and Model III: Community Care - mothers in community adherence clubs and infants in well-baby clinics) in South Africa, from the provider and patient's perspectives. Costs are presented in 2019 United States Dollars (US $). Results Lifelong ART can be considered cost-effective in Eswatini with an ICER of US $984 per mother retained in care to six months postpartum. In Cape Town, South Africa, Routine Care cost US $226 per mother-infant pair per annum; Integrated Care cost US $341; and Community Care cost US $254. Annual patient costs (direct and indirect costs) for Models I-III, were US $30-55, US $23-45 and US $76 per mother-infant pair respectively. Comparatively Community Care was the most cost-effective model with an ICER of US $97 per mother-infant pair retained and mother virally suppressed. Scaling-up Community Care nationally in South Africa would require US $5 720 096 more than Routine Care, 0.2% of the total health budget for 2020/21. Conclusions This work has generated novel empirical data in the form of new cost estimates and cost comparisons across different models of care. It has also provided a unique comparison of the different models of care using a cost-effectiveness analysis; and further a novel budget impact analysis of different approaches to rolling these strategies out. This data has helped to fill the gap in the evidence base for instance lifelong ART was implemented in Eswatini as a direct result of the Safe Generations study findings. Community Care was found to be cost-effective and if scaled up nationally in South Africa would only require a small increment of the total health budget. However, we recommend a mixture of models of care to cater for the needs and preferences of patients. Decision makers can use the empirical findings to help set realistic budgets in Southern Africa and explore ideal model implementation to support mother-infant pairs in the crucial postpartum phase