Sample Size Reestimation In Clinical Trials By Conditional Power Approach

Randomized controlled trials (RCT) in the1960s had emerged as the gold standard for evaluation of new drug in medical research. At the trial design stage, some crucial information used to calculate the sample size may not be available, or may be available but with a certain degree of uncertainty. It is necessary to check the validity of those assumptions for sample size calculating using interim data from the study and make sample size adjustment if necessary. Conditional power (CP) is the probability that the final study result will be statistically significant, given the data observed thus far and a specific assumption about the data to be observed in the remainder of the study. Brownian motion can be used to describe the distribution of the interim Z-test value, the corresponding B-value, and the CP values under a specific assumption about the future data. Based on Lan and Trost’s method (1997), this article proposed the condition that stop the trial early to claim significance and demonstrated that if the interim decision boundaries are proposed appropriately, the alpha inflation, due to sample size re-estimation and early stopping to claim significance, could be balanced with alpha deflation from early stopping for futility by the stochastic simulation method. In addition, the procedures of study design and practical significance are interpret by a real case.