Defining and evaluating novel procedures for involving patients in Core Outcome Set research: creating a meaningful long list of candidate outcome domains.

Plain English summary: Outcome domains are aspects of a condition that matter to patients and clinicians and can be measured to assess treatment effects. For tinnitus, examples include 'tinnitus loudness' and 'ability to concentrate'. This study focuses on the first stage of agreeing which outcome domains should be measured in all clinical trials of tinnitus. Crucially, it involves identifying outcome domains, prior to a voting process. This article describes how we effectively involved patients in that study design process, and reflects on the impact of their input.The study first compiled a long list of all possible outcome domains before asking interested parties, including patients, to vote which ones to include. Ensuring patients fully participate in this process holds unique challenges as it can be long, repetitive and its purpose far removed from their needs. These challenges may be addressed by involving patients in designing the research. There is evidence that other research teams are doing this, but its reporting is not detailed enough to guide others. Our paper seeks to address this.We describe how we involved patients (people living with tinnitus) in creating a long list of outcome domains that we included in our study. We also reflect on the benefits this brought. Two patients partnered with us in designing the survey. We also consulted an independent patient review panel. Involving patients reduced the list of domains included in the survey and made domain names and associated descriptions clearer. Our resulting survey performed well in recruiting and retaining patients as participants.

Abstract: Background Tinnitus is a complex audiological condition affecting many different domains of everyday life. Clinical trials of tinnitus interventions measure and report those outcome domains inconsistently and this hinders direct comparison between study findings. To address this problem, an ongoing project is developing a Core Outcome Set; an agreed list of outcome domains to be measured and reported in all future trials. Part of this project uses a consensus methodology ('Delphi' survey), whereby all relevant stakeholders identify important and critical outcome domains from a long list of candidates. This article addresses a gap in the patient involvement literature by describing and reflecting on our involvement of patients to create a meaningful long list of candidate outcome domains. Methods Two Public Research Partners with lived experience of tinnitus reviewed an initial list of 124 outcome domains over two face-to-face workshops. With the Study Management Team, they interpreted each candidate outcome domain and generated a plain language description. Following this, the domain names and descriptions underwent an additional lay review by 14 patients and 5 clinical experts, via an online survey platform. Results Insights gained from the workshops and survey feedback prompted substantial, unforeseen modifications to the long list. These included the reduction of the number of outcome domains (from 124 to 66) via the exclusion of broad concepts and consolidation of equivalent domains or domains outside the scope of the study. Reviewers also applied their lived experience of tinnitus to bring clarity and relevance to domain names and plain language descriptions. Four impacts on the Delphi survey were observed: recruitment exceeded the target by 171%, there were equivalent numbers of patient and professional participants ( n  = 358 and n  = 312, respectively), feedback was mostly positive, and retention was high (87%). Conclusions Patient involvement was an integral and transformative step of the study design process. Patient involvement was impactful because the online Delphi survey was successful in recruiting and retaining participants, and there were many comments about a positive participatory experience. Seven general methodological features are highlighted which fit with general principles of good patient involvement. These can benefit other Core Outcome Set developers.