Swaziland has one of the highest adult HIV prevalence rates (26%) worldwide. As Swaziland and many other countries in sub-Saharan Africa are moving to universal test-and-treat (UTT), or Early Access to ART for All (EAAA), policies, it is critical to understand the cost implications of this change to inform long-term financial planning. The study presented here is one of the first empirical costing study of EAAA in a public-sector health setting. We compare average patient costs under standard-of-care (SOC) vs. under EAAA.
From September 2014 through July 2016, we collected comprehensive data on facility-level ART costs from 7 facilities prior to and after transitioning to EAAA as part of a large-scale randomized stepped-wedge health systems trial of EAAA. We used a ‘bottom-up costing’ approach extracting data from facility budgets, expenditure reports, and patient records. SOC ART eligibility threshold was ≤350 cells/ml until December 2015, when it changed to ≤500 cells/ml. We used an activity-based allocation method for costs shared with non-ART patients. The costs include medications, laboratory services, direct and indirect personnel, equipment and administrative services. Total costs were divided by the total number of ART patient-years over the period the facility experienced SOC and EAAA to derive ART cost per patient per year (PPPY).
The average facility-level cost for ART patients was $277 PPPY (95%CI: 184-371) in SOC compared to $254 PPPY (95%CI: 209-298) in EAAA (p=0.66). The cost of ARVs was $113 PPPY (95%CI: 111-116) in SOC and $110 PPPY (95%CI: 108-112) in EAAA (p=0.04). The personnel costs were $130 PPPY (95%CI: 43-217) in SOC and $96 PPPY (95%CI: 51-141) in EAAA (p=0.51). Laboratory costs were $22 PPPY (95%CI: 11-33) in SOC and $37 PPPY (95%CI: 29-44) in EAAA (p=0.05).
We present the first direct comparison of public-sector patient costs under EAAA and SOC. Even though we would expect ART patients under EAAA to be on average healthier than patients under SOC, average public-sector costs per ART patient are essentially the same under the two treatment policies. Differences in funding requirements for SOC vs. EAAA will thus be largely driven by the number of patients receiving treatment. The larger average laboratory costs in EAAA were explained by increased uptake of viral load monitoring.