Robustness of Average Power for the Design in Clinical Trials

Robustness of Average Power for the Design in Clinical Trials

陳奕錩(Yi-Chang Chen)

指導教授 ： 黃郁芬 ； 蕭金福

國立中正大學/理學院/數學系統計科學研究所/碩士(2012年)

並列摘要

In conventional clinical trials, we usually consider the sample size to achieve a desired power for a specified treatment effect. In practice, there may be considerable uncertainty in the real treatment effect. Therefore, power might not provide a precise probability to calculate the sample size. That is, power often gives an over-optimistic estimate.

We argue that average power is an important measure in clinical trials. In fact, average power is the expect power, averaged over the prior distribution for the unknown true treatment effect. We feel that using the average power instead of the conventional of power is more appropriate to calculate the sample size. So, choosing a good prior is very important. In this thesis, we provide some common priors, and want to see the robustness of the average power with respect to changes in prior.

並列關鍵字

Average power ； Bayesian ； Bayesian analysis ； Power ； Sample size

參考文獻

Chuang-Stein, C.(2006). Sample size and probability of a successful trial.

Pharmaceutical Statistics; 5:305-309.

O’Hagan, A. and Stevens, J.W. (2001). Bayesian assessment of sample size for clinical trials of cost-effectiveness. Medical Decision Making; 21:219–230.

Spiegelhalter, D.J. and Freedman, L.S.(1986). A predictive approach to selecting the size of a clinical trial based on subjective clinical opinion. Statistics in Medicine; 5:1–13.

Carlin, B. P. and Louis, T. A. (1996). Identifying prior distributions that produce specific decisions, with application to monitoring clinical trials, in Bayesian Analysis of statistics and Econometric (Berry, D. A.; Chaloner, K. M.; Geweke, J. K. eds.), John Wiley & Sons, New York.