CONFIDENCE-INTERVAL CALCULATION FOR A SEQUENTIAL CLINICAL-TRIAL OF BINARY RESPONSES

Consider a sequential test applied to two streams of binary responses in a comparative clinical trial. After the completion of such a trial Woodroofe's (1992) technique for accurate confidence interval estimati on for a treatment difference can be applied. In this paper the techni que is extended to provide confidence intervals for the individual abs olute success probabilities of the two treatments in the trial, for wh ich no current methodology exists. Accuracy is explored by simulation of coverage probabilities and individual confidence limit probabilitie s. The simulations concern a triangular test and an O'Brien & Fleming test.