Assessing efficiency and costs of scaling up HIV treatment.

INTRODUCTION Whereas cost-effectiveness/utility analyses theoretically assess efficiency in HIV treatment, in practice they are of limited use to policy makers who are also concerned with the total costs of scaling up. This paper proposes an approach to simultaneously assessing both factors when setting priorities for HIV treatment. METHODS Three interventions were assessed: a no antiretroviral therapy (ART) status quo, ART including first-line only, and ART including first and second-line regimens. Data were from a cohort receiving healthcare in a poor South African setting. Markov modelling was used to calculate patient-level lifetime costs and quality-adjusted life-years (QALY) as well as population-level total costs and QALY in each intervention. Linear programming was used to assess efficiency at the population level. RESULTS First-line ART costs US$795 per QALY gained compared to no ART, while first and second-line costs US$1625 compared to first-line alone. The efficiency of either ART strategy depends on the HIV treatment budget. If this is less than US$10 billion during the planning period, first-line ART is most efficient. A combination of first-line with first and second-line treatment is most efficient if the budget is US$10-12 billion. Using both first and second-line treatment for everyone becomes efficient as the main strategy only at budgets greater than US$13 billion. CONCLUSION An approach has been developed to HIV treatment priority setting that simultaneously considers efficiency and the costs of scaling up. This can help to establish explicit and evidence-based priorities and budgets to meet scaling up challenges.