A new program to compute and evaluate continuously monitored stopping boundaries for clinical trials.

We present code for the calculation and evaluation of continuously monitored stopping boundaries for use in one-arm and two-arm clinical trials. These designs were first developed for one-arm trials by Thall, Simon and Estey (TSE) (P.F. Thall, R. Simon, E.H. Estey, Bayesian sequential monitoring designs for single-arm clinical trials with multiple outcomes, Stat. Med. 14 (1995) 357-379). Our code corrects some problems in the original TSE algorithms and extends these algorithms for use in a two-arm trial setting. It is written in S-Plus to improve interactivity for the statistically adept user, and employs external routines, dynamically loaded into S-Plus, to improve calculation efficiency. Efficient versions of our code require both a C compiler and the S-Plus program. Our code has been tested in UNIX and Microsoft Windows environments, and compiled code is available from our website. A numerical integration routine for the convolution of beta distributions is included.