INTERPRETATION AND ESTIMATION OF CONDITIONAL POWER IN INTERIM ANALYSES OF CLINICAL TRIALS: OFTEN IGNORED IN DECISION MAKING PROCESS

Abstract

In group sequential trials, the interim results are more promising in the early termination of a trial either for efcacy or futility of the trial. This reduces the cost and time implications. Moreover, interim analyses play a key role to tackle the problem that arises due to adverse effects. In concern with the early stopping of trials, there are numerous stopping methods among them the Conditional power (CP) approach is best recommended. The CP approach provides the probability of getting signicant results at the end of the study given the data observed so far. There are very few Indian studies that had incorporated the concept of CP and made decisions based on the results. This study popularises the CP approach detailing computing and its interpretation and is thereby facilitating clinical researchers to use this approach effectively. We have used real-time and hypothetical examples and illustrated the concept of CP under trend, CP under null, and CP under design. We suggest stopping the study for efcacy when the CP under null is greater than 80% and stopping the study for futility when CP under trend is less than 30%.