A systematic review of core outcome set development studies demonstrates difficulties in defining unique outcomes

Abstract Objectives Core Outcome Set (COS) development often begins with a systematic review to identify outcomes. Reviews frequently show heterogeneity in numbers of outcomes reported across trials. Contributing to this is a lack of a uniform definition for an outcome. This study proposes a first working definition for a unique trial outcome to support reporting a quantitative assessment of outcome reporting heterogeneity (ORH). Study Design and Setting Eligible COS literature, (development papers, protocols and reviews) were identified using the COMET database, Ovid MEDLINE and PubMed. Outcome numbers, definitions, timing and grouping methodology were examined. Results 132 studies were included. 82(88.1%) studies (excluding protocols) reported a total number of unique outcomes (median 82, range 12-5776, IQR 261). Timing of assessment was reported in 32(31.4%) studies. Methods to group similar outcomes were reported in 8 (7.8%) papers. No study defined how outcomes were agreed as different and how final numbers of unique outcomes were determined. It is proposed that a unique outcome requires original meaning and context. Thus ORH is suggested to be the reporting of multiple unique outcomes across trials related to one healthcare condition. Conclusion This review identified inconsistencies in how authors define, extract, group and count trial outcomes. Further work is needed to refine our proposed definitions to optimise COS development and allow a quantifiable measure of ORH.