Commentary: Why we should report results from clinical trial pilot studies

Should clinical trial pilot studies be reported? Thebane, Ma, Chu, et al. argue in their discussion of the conduct of pilot studies, that they should be published [1]. Others, such as the editorialist for The Diabetes Educator, take a more cautious approach to publishing pilot studies [2]. I agree with those who advocate publication. I might also note that a September 27, 2012 search of “randomized clinical trial pilot studies” in PubMed contained over 10,400 citations. Although some of these may not reflect individual pilot studies, there may, though, be many more (without “randomized” in the search term, the number is over 22,000). First, I will define my terms. By a pilot study, I mean one that is conducted preliminary to a full-scale late phase (phase 3 or 4, in common terminology) trial. The pilot study is conducted in order to address questions important for the design and conduct of the full-scale trial, and is sometimes called a feasibility or vanguard study. Obviously, the usual early phase (phase 1 and 2) studies could serve those purposes. (The results of these studies are often not reported, and whether they should be is another issue.) But, often, even after the early phase studies are done, we still have questions about participant screening and enrollment, adherence over some period of time, administration of the intervention in community settings, outcome assessment, and the like. Pilot studies are designed to answer those sorts of questions that are essential in the development of the full-scale trial. Pilot studies can also be of two sorts. One is an “internal pilot,” where, if it is successful and meaningful protocol changes (a phrase that is subject to judgment) are not made in the transition to a full-scale trial, it more or less seamlessly merges into the larger, definitive clinical trial. In that situation, the full-scale trial rules about reporting interim data would apply. An example is the pilot study that was done prior to the full-scale Antiarrhythmics Versus Implantable Defibrillators (AVID)