Guidance for defining outcomes in clinical trials

Defining outcomes completely before conducting clinical trials helps to mitigate concerns about reporting biases; however, there is limited guidance to help investigators define outcomes completely.

Building on a previous framework, we developed guidance for defining trial outcomes completely and consistently. We reviewed literature, developed preliminary guidance, and refined guidance iteratively through pilot testing. We randomly selected 225 randomized controlled trials (RCTs) from ClinicalTrials.gov, which evaluated 3,424 outcomes. Two raters followed a structured approach to define each outcome. When raters encountered outcomes they could not define, we refined our guidance. We continued this process until no further changes were needed. We finalized the guidance through multiple drafts and discussions among authors.

In clinical trials, a completely defined outcome includes six elements: outcome domain, specific measurement, specific metric, cutoff (if applicable), variable type, and time point. We developed guidance for naming specific measurements for both subjective and objective outcomes. We expanded on prior work by describing more comprehensive categories for specific metrics. To be consistent with the estimand framework, we replaced "method of aggregation" with two distinct elements that describe values in individuals (rather than summary measures): "variable type" and "cutoff." We further defined "cutoff" with three sub-elements, and we defined "variable type" as whether the value for each individual is continuous or categorical.

Trialists and sponsors could use this framework and guidance to define outcomes in trial registrations, protocols, and results reports. When all six elements are specified, readers can better determine whether reported outcomes were prespecified and whether different trials assessed the same outcomes.