Abstract:Background COVID-19 vaccines were rolled out in South Africa beginning in February 2021. In this study we retrospectively assessed the cost-effectiveness of the vaccination programme in its first two years of implementation. Method We modelled the costs, expressed in 2021 US$, and health outcomes of the COVID-19 vaccination programme compared to a no vaccination programme scenario. The study was conducted from a public payer's perspective over two time-horizons – nine months (February to November 2021) and twenty-four months (February 2021 to January 2023). Health outcomes were estimated from a disease transmission model parameterised with data on COVID-19-related hospitalisations and deaths and were converted to disability adjusted life years (DALYs). Deterministic and probabilistic sensitivity analyses (DSA and PSA) were conducted to assess parameter uncertainty. Results Incremental cost-effectiveness ratio (ICER) was estimated at US 1300 per DALY averted. When 85% of all excess deaths during these periods were included in the analysis, ICERs in the first and second study periods were estimated at US 660 per DALY averted, respectively. In the PSA, almost 100% of simulations fell below the estimated opportunity cost-based cost-effectiveness threshold for South Africa (US$2300 DALYs averted). COVID-19 vaccination programme cost per dose had the greatest impact on the ICERs. Conclusion Our findings suggest that South Africa's COVID-19 vaccination programme represented good value for money in the first two years of rollout.

A cost-effectiveness analysis of South Africa's COVID-19 vaccination programme