|
Activity Number:
|
263
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 10, 2004 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #301646 |
|
Title:
|
Cost-utility Analysis of HIV Prevention Interventions: Estimates and Uncertainties
|
|
Author(s):
|
Matthew J. Hayat*+ and Prakash Laud and Raymond G. Hoffmann and Ana Johnson-Masotti
|
|
Companies:
|
Emmes Corporation and Medical College of Wisconsin and Medical College of Wisconsin and McMaster University
|
|
Address:
|
NIDCD, 6120 Executive Blvd., MSC 7180, Bethesda, MD, 20892,
|
|
Keywords:
|
HIV ; cost ; quality of life ; Bayesian ; effectiveness ; utility
|
|
Abstract:
|
Cost-utility analyses provide useful guidance to policymakers in comparing different treatments for a disease or in allocating societal resources to alleviating the ill effects of different diseases. We first describe a randomized study conducted to assess the effectiveness of three different strategies designed to increase protective behavior in mentally ill adults. Next we give an overview of statistical measures used in cost-utility analyses, including a discussion of some related pitfalls. A further complication arises when costs or utilities cannot be measured directly, as in the case of HIV prevention programs. Observable behavioral changes are translated via models into QALYs (quality adjusted life years) saved by a prevention program. These models contain parameters about which only partial information is available. We consider inference models that combine this information with an accounting of the sampling variability in the behavioral changes. Resulting methods are illustrated on data from the above mentioned HIV prevention study.
|
