The Use of Auxiliary Information in the Design of a Clinical Trial.
HARVARD UNIV CAMBRIDGE MASS DEPT OF STATISTICS
Pagination or Media Count:
Various strategies are investigated for the incorporation of auxiliary information into the design of a clinical trial to compare two treatments. Out of a total of N patients with the disease in question, a fraction 2P of them is to be placed in a clinical trial comparing the two treatments, with the winning treatment to be given to the remaining N1-2P patients. The direct information from the trial is combined with information derived from various outside sources in order to make the final decision. Strategies investigated include simple pooling, reliance on the auxiliary information alone, the theoretically optimal Bayes strategy, and a 2-stage plan whereby the decisiveness of the auxiliary information is evaluated in stage 1. Where applicable, optimal values of P are found. The 2-stage plan is found to be the overall best strategy. Conditions are given under which the Bayes strategy and the pooling strategy have approximately the same efficiency. The robustness of the latter two strategies with respect to knowledge of certain population parameters is also investigated. Modified author abstract
- Medicine and Medical Research
- Statistics and Probability