Group sequential methods for clinical trials allowing early acceptance of Ho and incorporating costs

When interim accumulated results from a clinical trial clearly indicate that the null hypothesis should be ‘accepted’ or ‘rejected’, ethical and cost considerations may dictate terminating the trial and making the appropriate decision. Pocock (1977) has described a group sequential procedure for clinical trials that permits early termination only for rejecting the null hypothesis. This paper extends Pocock's procedure to allow early termination with acceptance of the null hypothesis. Theoretical and simulation results demonstrate that designs allowing early acceptance of the null hypothesis can have lower expected costs than designs that do not, even when the alternative hypothesis is true.