A sequential trial of pain killers in arthritis: issues of multiple comparisons with control and of interval-censored survival data.

A large clinical comparison of pain killers used in the treatment of arthritis was conducted using a sequential design. The trial presented two major challenges: there were three treatments to be compared, and patients were assessed annually for up to 7 years. Treatment comparisons were made by selecting appropriate pairwise comparisons, to which standard theory could be applied. Rules concerning actions to take in the event of each stopping criterion being reached were evaluated by simulation. The annual assessments were used to create an interval-censored survival time: the year during which "disease progression" first occurred.