Underpowered Trials at Trial Start and Informed Consent: Action Is Needed Beyond the COVID-19 Pandemic

To be ethical, clinical trials must (among other things) be scientifically valid and must yield socially or scientifically valuable information from reliable data.1 These requirements are closely related but not identical. A poorly designed study cannot yield valuable information, but a well-designed study can yield trivial or commonplace information of no social or scientific value. When deciding upon the design, one should first consider whether the trial will be controlled: only in special circumstances (e.g. ultra rare diseases, early clinical development drug trials in oncology) could single group trials provide limited valid information. In controlled trials, measures to prevent bias (e.g. randomisation, masking) are critical to ensure scientific validity. In pilot (or feasibility) trials, where the results will allow, among other things, the estimation of the sample size of a subsequent large trial, the number of participants is usually limited. Other types of trials must enrol a sufficient number of expected participants. The sample size of the trial determines its ability to demonstrate the effect size difference of the interventions if it exists. Lacking scientific validity that renders trials lacking social value is a common source of waste in clinical research, diverting participants and human and technical resources from conducting other properly designed trials.