We used a futility stopping boundary, where we stopped in case of a negative trend, i.e. the observed interim estimate of the treatment group is worse than of the control group.
← all excerpts
Randomization in clinical trials with small sample sizes using group sequential designs.
1
—
—