HEALTH UTILITY ESTIMATES AMONG PERSONS LIVING WITH HIV IN THE U.S. – IMPLICATIONS FOR COST-EFFECTIVENESS MODELING AND FUTURE RESEARCH NEEDS

   Purpose: Cost-utility analyses (CUA) use quality-adjusted life-years (QALYs) to determine optimal prevention and clinical strategies. However, published health utility estimates specific to persons living with HIV in the United States are limited and increasingly outdated as HIV has evolved into a chronic rather than acute disease due to highly effective antiretroviral therapy (ART). We examine published U.S. HIV-related preference-based health utility (HU) estimates and those derived from health-related quality-of-life (HR-QOL) assessments, present previously unreported data from 2 additional studies, and discuss implications for CUA.

   Methods: U.S. HIV modeling studies referenced 6 HU and HR-QOL sources (see Table 1). Previously unreported HR-QOL estimates are presented from the INSIGHT Strategies for Management of Anti-Retroviral Therapy (SMART) trial and CDC's Medical Monitoring Project (MMP). SMART was a randomized trial comparing continuous and intermittent CD4 count guided ART with SF-12 HR-QOL assessment during clinic visits for U.S. participants. MMP is an annual cross-sectional survey of persons living with HIV receiving medical care in the U.S. that included SF-12 assessment in 2007. To approximate the methodologically preferred HU approach, we converted SMART baseline and MMP HR-QOL data to SF-6D estimates. For each of the 8 sources examined, we compared the data collection time frame, assessment method, and key estimates.

   Results: Data collection for all 6 previously published estimates pre-date 1998 and the widespread use of ART. Two of the 6 published sources provided direct HU estimates, while all others were based on HR-QOL data. Estimates derived from SMART and MMP SF-12 data were markedly similar across key stages of HIV disease (0.69-0.77), and differed substantially from those reported in the most frequently cited, HU, source (0.70-0.94, Tengs et al.).

   Conclusion: As the context of HIV has significantly changed over the past two decades and the majority of the studies in the literature do not reflect methodologic best practices (i.e. direct preference-based HU estimation), updated utilities are needed. To best inform health policy, CUA using currently available estimates should present results for both QALYs and life years.