In reply: Should we ever stop clinical trials for efficacy?
- 128 Downloads
To the Editor,
We thank Mr. Fergusson and Dr. Griesdale for presenting a well-reasoned discussion of the consequences of stopping a trial such as ours early.1,2 We generally agree that effect overestimation and lack of precision are unfortunate by-products of stopping a trial early. Nevertheless, we disagree with their suggestion that clinical trials should be stopped only “when it is obvious that clinical equipoise no longer exists.” The existence of different stopping rules indicates that deciding whether to stop or continue a trial is neither straightforward nor trivial.3 While P values are what lie on the surface, there are ethical reckonings involved. It requires weighing the current patients’ good against the potential benefit to society.4 Each physician is required to balance the duty of care with the search for truth.5 Each patient desires the most effective treatment for themselves, not to serve an abstract scientific goal. To reconcile these interests, the power to stop or continue is rightfully vested with an institutional review board (IRB).
Though IRBs follow the principle of equipoise, this equipoise depends on what evidence is available at a given time. The editorial by Andrade (and the majority of its supporting references) summarizing the inconsistent results in ketamine trials was unavailable during our interim analysis.6 Instead, the evidence with respect to our primary and many important secondary clinical outcomes significantly favoured ketamine over propofol. Together with findings from a paper published around the same time, our results may have led the IRB to conclude that equipoise no longer held.7
A larger sample number could give results that show either a greater or a lesser effect size, and would likely be a better estimate of the true treatment effect of our own study intervention. Nevertheless, IRBs are not required to stop or continue trials based on effect size, but rather on whether there is an effect at all.7 We are aware that our trial’s small size and single-site nature are unlikely to change practice. We do believe that our results contribute to the scientific evidence and warrant a larger multi-centre trial.
This research was conducted without funding.
Conflict of interest
No author has any commercial or other affiliations that are, or may be perceived to be, a conflict of interest.
This submission was handled by Dr. Hilary P. Grocott, Editor-in-Chief, Canadian Journal of Anesthesia.