Reconsidering registration requirements for trials randomising health care providers.

Abstract

Clinical trial registration plays a vital role in enhancing transparency, reducing reporting bias, and upholding the integrity of the evidence used in clinical and policy decision-making. The International Committee of Medical Journal Editors (ICMJE) has produced a set of recommendations covering trial registration for its member journals, but they are widely adopted by many journals. It recommends that trials are registered prospectively, and that trial registration is clearly reported in all related manuscripts. However, ICMJE's recommendations are ambiguous when it comes to the registration of clinical trials involving healthcare providers (HCPs), indicating that if trials study the effect of an intervention only on HCPs (e.g. provider knowledge or attitudes), then registration is not necessary. In this commentary, we argue that all clinical trials involving HCPs should be registered for three reasons. First, these trials (like any other research) are at an increased risk of reporting bias should they not be registered prospectively. Second, when HCPs are research participants, they are deserving of the same ethical protections as other types of research participants (and this includes trial registration). Finally, not exempting these trials from registrations would bring ICMJE in line with other international guidelines for health research (which do not exempt them). We conclude by advocating for a revision of the ICMJE trial-registration recommendation.