Current controversies in data monitoring for clinical trials.

This article presents some real-life challenges faced by clinical trial Data Monitoring Committees (DMCs), with the aim of clarifying some of the controversial issues that relate to both statistical stopping boundaries and DMC decision-making. Specific attention is given to what constitutes a sensible statistical boundary for stopping a trial early for benefit, bearing in mind that one usually needs proof beyond reasonable doubt of a treatment benefit sufficient to alter future clinical practice. Appropriate choices of stopping boundary for harm and futility are also discussed. The examples serve to illustrate that the practicalities of DMC decision-making require wise judgements based on a totality of evidence, making any statistical boundary just an objective guideline rather than a definitive stopping rule.