Selecting and collecting outcomes for clinical trials: an international qualitative study exploring stakeholder perspectives

Abstract Introduction Selecting and collecting data to support appropriate primary and secondary outcomes is a critical step in designing trials that can change clinical practice. In this study, we aimed to investigate who contributes to the process of selecting and collecting trial outcomes, and how these people are involved. This work serves two main purposes: 1) it provides the trials community with evidence to demonstrate how outcomes are currently selected and collected, and 2) it allows people involved in trial design and conduct to pick apart these processes to consider how efficiencies and improvements can be made. Methods One-with-one semi-structured interviews, supported by a topic guide to ensure coverage of key content. The Framework approach was used for thematic analysis of data, and themes were linked through constant comparison of data both within and across stakeholder groups. Participants were twenty-nine international trialists from various stakeholder groups, working primarily on designing and/or delivering phase III pragmatic effectiveness trials. Their experience spanned various funders, trial settings, clinical specialties, intervention types, and participant populations. Results We identified three descriptive themes encompassing the process of primary and secondary outcome selection, primary and secondary outcome collection, and the publication of outcome data. Within these themes, participants raised issues around the following: 1) Outcome selection: clarity of the research question; confidence in selecting trial outcomes and how confidence decreases with increased experience; interplay between different stakeholders; how patients and the public are involved in outcome selection; perceived impact of poor outcome selection including poor recruitment and/or retention; and use of core outcome sets. 2) Outcome collection: disconnect between decisions made by outcome selectors and the practical work done by outcome collectors; potential impact of outcome measures on trial participants; potential impact on trial staff workload; and use of routinely collected data. 3) Publication of outcome data: difficulties in finding time to write and revise manuscripts for publication due to time and funding constraints. Participants overwhelmingly focused on the process of outcome selection, a topic they talked about unprompted. When prompted, participants do discuss outcome collection, but poor communication between selectors and collectors at the trial design stage means that outcome selection is rarely linked with the data collection workload it generates. Discussion Stakeholders involved in the design and conduct of trials fail to connect decisions around outcome selection with data collection workload. Publication of outcome data and effective dissemination of trial results are hindered due to the project-based culture of some academic research.