Regression analysis on the variation in efficiency frontiers for prevention stage of HIV/AIDS

Abstract

Objectives:

To investigate how the cost effectiveness of preventing HIV/AIDS varies across possible efficiency frontiers (EFs) by taking into account potentially relevant external factors, such as prevention stage, and how the EFs can be characterized using regression analysis given uncertainty of the QALY-cost estimates.

Methods:

We reviewed cost-effectiveness estimates for the prevention and treatment of HIV/AIDS published from 2002–2007 and catalogued in the Tufts Medical Center Cost-Effectiveness Analysis (CEA) Registry. We constructed efficiency frontier (EF) curves by plotting QALYs against costs, using methods used by the Institute for Quality and Efficiency in Health Care (IQWiG) in Germany. We stratified the QALY-cost ratios by prevention stage, country of study, and payer perspective, and estimated EF equations using log and square-root models.

Results:

A total of 53 QALY-cost ratios were identified for HIV/AIDS in the Tufts CEA Registry. Plotted ratios stratified by prevention stage were visually grouped into a cluster consisting of primary/secondary prevention measures and a cluster consisting of tertiary measures. Correlation coefficients for each cluster were statistically significant. For each cluster, we derived two EF equations – one based on the log model, and one based on the square-root model.

Discussion:

Our findings indicate that stratification of HIV/AIDS interventions by prevention stage can yield distinct EFs, and that the correlation and regression analyses are useful for parametrically characterizing EF equations. Our study has certain limitations, such as the small number of included articles and the potential for study populations to be non-representative of countries of interest. Nonetheless, our approach could help develop a deeper appreciation of cost effectiveness beyond the deterministic approach developed by IQWiG.

Key words::

• Cost effectiveness
• Efficiency-frontier
• HIV/AIDS
• IQWiG
• Prevention

Transparency

Declaration of funding

The authors have no support or funding to report.

Declaration of financial/other relationships

Division of Medical Statistics, Graduate School of Medicine, Kobe University, Japan (MSK); The Center for the Evaluation of Value and Risk in Health, Institute for Clinical Research and Health Policy Studies, Tufts Medical Center, Boston, MA, United States (MSK, JTC, PJN); Graduate School of Health Management, Keio University, Japan (IK).

Data in this paper were presented at the 14th Annual Meeting of the International Society for Pharmacoeconomics and Outcomes Research, May 14–20, 2009, Orlando, Florida.