Abstract:
|
In many disease areas, when the efficacy outcomes involve multiple major events of different types or recurrent events of the same type, evaluation of all events might render a more sensitive measure of treatment effect provided major clinical efficacy could be so captured. In clinical trials in which complex clinical questions are of primary interest, the efficacy of a new treatment may be detected with greater statistical sensitivity from multiple events than from the first event alone. Both the cumulative rate function and the occurrence rate function may be adopted to study the treatment effect. Design and power consideration lends itself to justification of the statistical utility of multivariate failure-time analysis method. In this talk, treatment effectiveness in double-blind, randomized, controlled clinical trials will be evaluated under the typical assumption of non-informative censoring via a case example. In addition, the possibility that the censoring is informative will be explored in this example.
|