Abstract Background and Aims Patient-centered outcomes play an important role when assessing treatment effects. Patient-reported outcome (PRO) measures assessing symptoms and impacts are important tools to capture the patient perspective in clinical trials. However, careful consideration is needed when choosing PROs and PRO measures. Aspects such as the conceptual model, measurement model, validity, reliability, and responsiveness, as well as translation, interpretability of scores and burden for participants and investigators have to be considered when developing a patient-centered outcome strategy. The CONVINCE randomized controlled trial was funded by the European Union (HORIZON2020) to assess effects of high-dose hemodiafiltration (HDF) versus high-flux hemodialysis (HD). The primary outcome is all-cause mortality and one of the secondary outcomes is the experienced health status of the patients as measured by patient-reported outcomes. The aim of this study is to describe the approach for the development of a patient-centered outcome strategy using a domain-based approach that was implemented in the CONVINCE trial. Method The development of the patient-centered outcome strategy for the CONVINCE trial followed three consecutive steps: (1) targeted literature review to identify core outcome sets, as well as concepts and PRO measures in the field of dialysis; (2) semi-structured interviews and focus groups with dialysis patients (n = 13) and clinicians (n = 4) to determine health aspects, i.e. symptoms and impacts most relevant to patient population; (3) triangulation of results to determine the most relevant health domains to assess within the context of the CONVINCE trial and determination of PRO measures in due consideration of psychometric properties of respective instruments. Results We identified 97 PRO measures reported in previous dialysis studies, assessing more than 11 health domains and 41 disease-specific symptoms/problems relevant for dialysis patients. Core outcome sets, such as the SONG-HD and the ICHOM standard set for chronic kidney disease, indicated in total 19 outcomes that we classified as potential PROs. Interviews with patients and clinicians confirmed health domains and outcomes that were identified in the literature and provided further insights in terms of the level of importance of health domains. To balance the comprehensiveness of assessments and respondent burden, we decided to concentrate on 9 domains and 20 symptoms/problems according to the objectives of the trial. For the more proximal outcomes, i.e. recovery time and recovery time pattern, new items were developed by the CONVINCE consortium applying established state-of-the-art methods. For generic outcomes, e.g. fatigue, physical function, we decided to apply psychometrically advanced tools from the PROMIS® initiative, validated within a general population as well as various chronic health conditions. Conclusion Building upon core outcome sets by confirming health concepts most relevant for dialysis patients within the specific context of the CONVINCE trial, enabled us to develop a robust patient-centered outcome measurement strategy for the CONVINCE trial that assesses relevant health domains of dialyses patients in granular fashion using a domain-based approach. Further validation of the PRO measures should be considered, especially establishing meaningful change estimates to support interpretation of change in PRO measurement scores. Conference Abstract ERA EDTA; May 23-26, 2024; Stockholm, Sweden Kathrin I. Fischer, Peter J. Blankestijn, Krister Cromm, Michiel L. Bots, Claudia Barth, Bernard Canaud, Jörgen Hegbrant, Giovanni Strippoli, Andrew Davenport, Mark Woodward, Matthias Rose for the CONVINCE Scientific Committee and CONVINCE Investigators
Read full abstract