Sequential Medical Trials: Some Comments on F. J. Anscombe's Paper

Abstract This is a reply to Professor Anscombe's review article of the author's book, “Sequential Medical Trials.” It is suggested that the use of likelihood inference may have misleading consequences. Anscombe's optimal sequential designs, while giving insight into the problem of designing clinical trials, would be difficult to put into practice and may be based on unacceptable ethical assumptions.