| Session Title: 'Expected Value of Perfect Information' in Health Economic Evaluation |
| Demonstration Session 519 to be held in Centennial Section H on Friday, Nov 7, 9:15 AM to 10:45 AM |
| Sponsored by the Costs, Effectiveness, Benefits, and Economics TIG |
| Presenter(s): |
| J Michael Menke, University of Arizona, menke@u.arizona.edu |
| Abstract: When do you have enough information or knowledge about a system, program, or treatment to make a sound decision? The expected value of perfect information (EVPI) when used with Bayesian methods in a decision analytic framework can address questions of whether resources should continue to be directed to system, program, or treatment effectiveness research. EVPI estimates the cost of uncertainty reduction in research and evaluation by indirectly estimating the cost of attaining 'perfect information', where the decision becomes error-free. This estimate is useful to inform health policy decisions, especially under resource constraints. Though effectiveness and cost effectiveness studies may help choose maximum health care or social welfare functions, they may still be poor guides, given the opportunity costs of overlooking alternative effective approaches. EVPI is a powerful tool for informing decisions by looking closely at uncertainty, and how much we are willing to spend to reduce it. |