The Estimation and Use of Confidence Intervals in Economic Analysis
- PDF / 523,738 Bytes
- 8 Pages / 504 x 720 pts Page_size
- 75 Downloads / 204 Views
Drug Infwmufion Jorrniol. Vol. 33. pp. 841-848. lW Rimed in che USA. All rights reserved.
Copyright 6 19% D N Information ~ Association h C .
THE ESTIMATION AND USE OF CONFIDENCE INTERVALS IN ECONOMIC ANALYSIS MOHANV. BALA,PHD Group Leader. Outcomes Research, Centocor. Inc., Malvern. Pennsylvania
JOSEPHINE MAUSKOPF,PHD Center for Eonomics Research, Research Triangle Institute, Research Triangle Park, North Carolina
This paper summarizes the diflerent methods used to estimate confidence intervals for cost-effectiveness ratios, along with their mnin strengths and weaknesses. We then examine the cost-effectiveness panel’s recornmendation on how to use these confidence intervals in decision mnkirrg. We conclude thirt the guidelines on the use of confidence intervals need to be based in decision theory and that further work is needed to guide the use of confidence intends in health care decision making. We propose Bayesian analysis as a promising theoretical ,framework for integrating uncertainty into health economic analysis. Kes Words: Cost effectiveness: Confidence interval
INTRODUCTION THE INCREMENTAL cost-effectiveness ratio (ICER) is a common metric used to compare alternative therapies in health economic analysis. ICER is the ratio of the difference in costs between two treatments to the difference in effectiveness. If one treatment is both more effective and less expensive, there is no need to compute this ratio since the choice is clear. If this is not the case, however, the ICER is compared to a threshold value to determine whether the gain in effectiveness is worth the additional expense. Most economic analyses present a point
Presented at the DIA 34th Annual Meeting “Thinking Globally: Product Development, Registration. and Marketing in the New Millennium.” June 7-1 l , 1998, Roston, Massachusetts. Reprint address: Mohan V. Bala. PhD, Group Leader. Outcomes Research, Centocor, Inc., 200 Great Valley Parkway, Malvern, PA 19355. E-mail: balm@ centocor.com.
estimate of the ICER based on the mean values of cost and effectiveness. The effectiveness (usually approximated by efficacy) estimates are commonly obtained from a clinical trial, while the cost estimates are obtained either directly from the trial, or from secondary data sources. It is important to note that there is uncertainty surrounding the point estimates of cost and effectiveness, which in turn results in uncertainty around the point estimate of the ICER. The traditional means of dealing with this uncertainty has been sensitivity analysis. This consists of varying the relevant efficacy and cost parameters within valid ranges and examining the impact of these changes on the ICER. This technique of dealing with uncertainty has several drawbacks as pointed out by Willan and O’Brien (1). There are no established standards to determine which parameters should be vaned and what the relevant range of variation should be. There are also no guidelines available to interpret
84 I
Downloaded from dij.sagepub.com at UNIV ARIZONA LIBRARY on May 31, 2015
Data Loading...
