Consider sequential trials comparing two treatments with binary responses. The goal is to derive accurate confidence sets for the treatment difference and the individual success probabilities of the two treatments. We shall begin with the signed-root transformation as a pivot and then apply the approximate theory of Weng and Woodroofe [11] to form accurate confidence sets of these parameters. The explicit correction terms of the pivots are obtained. The simulation studies agree well with the theoretical results.

[1] P. Armitage. Numerical studies in the sequential estimation of a binomial parameter. Biometrika, 45:1-15, 1958.
[2] I.V. Basawa and B. L. S. P. Rao, Stochastic Process. Academic Press, London, 1980.
[3] M. N. Chang. Confidence intervals for a normal mean following a group sequential test. Biometrics, 45:247-254, 1989.
[4] D. S. Coad and M. Woodroofe. Corrected confidence intervals after sequential testing with applications to survival analysis. Biometrika, 83:763-777, 1996.
[5] K. M. Facey and J. Whitehead. An improved approximation for calculation of confidence intervals after a sequential clinical trial. Statist. Med., 9:1277-1285, 1990.
[6] G. L. Rosner and A. A. Tsiatis. Exact confidence limits following group sequential test. Biometrika, 75:723-729, 1988.
[7] D. Siegmund. Estimation following sequential testing. Biometrika, 65:341-349, 1978.
[8] D. Siemund. Sequential Analysis. Springer, New York, 1985.
[9] S. Todd and J. Whitehead. Confidence interval calculation for a sequential clinical trial of binary responses. Biometrika, 84:737-743, 1997.
[10] S. Todd, J. Whitehead, and K. M. Facey. Point and interval estimation following a sequential clinical trial. Biometrika, 83:453-461, 1996.
[11] R. C. Weng and M. Woodroofe. Integrable expansions for posterior distributions for multiparameter exponential families with applications to sequential confidence levels. Statistica Sinica, 10:693-713, 2000.
[12] J. Whitehead. The Design and Analysis of Sequential Clinical Trials. Ellis Horwood, Chichester, 1983.
[13] M. Woodroofe. Very weak expansions for sequentially designed experiments: linear models. Ann. Statist., 17:1087-1102, 1989.
[14] M. Woodroofe. Estimation after sequential testing : A simple approach for a truncated sequential probability ratio test. Biometrika, 79:347-353, 1992.
[15] M. Woodroofe. Integrable expansions for posterior distributions for one-parameter exponential families. Statistica Sinica, 2:91-111, 1992.