Development of a Mortality Risk Prediction Tool for People With Kidney Failure: A Nominal Group Technique Study of End-User Priorities

Many mortality risk prediction tools for individuals with kidney failure are available; however, their development did not involve end-users during the design process, and none are widely utilized in clinical practice. We aimed to identify the needs and preferences of end-users to inform the development and enhance the usability of a mortality risk prediction tool for people with kidney failure. A half-day, online consensus workshop was conducted using a modified nominal group technique. People with lived experience of kidney failure (patients with or without receipt of kidney replacement therapy and their caregivers) or kidney failure management (healthcare providers and policymakers) recruited from across Canada. Preferences were elicited in three topic areas: the tool’s intended use, timing (prediction horizon and update frequency), and relevant predictor variables. Conventional content analysis of discussion transcripts was conducted to elaborate on the findings. Eighteen individuals from across five provinces participated in the workshop, including seven patients, three caregivers, and eight healthcare providers or policymakers. Participants prioritized the following: (1) tool use within clinics and in consultation with nephrologists; (2) personalization of the prediction time horizons and reassessment of risk prediction following changes in clinical condition; and (3) inclusion of co-existing conditions, kidney failure characteristics (e.g., unplanned dialysis start), and frailty status as key predictor variables. Analysis of transcripts identified several factors influencing the tool’s usability, including provider trust, comfort in discussing mortality risk, patient privacy, availability of follow-up and support systems, patient readiness, and feasibility of incorporating desired predictor variables. Media-based recruitment and participation barriers may have limited the representativeness of the patient sample. End-users of mortality prediction tools for people with kidney failure prioritized design and use considerations. Co-design of future tools to align with end-user preferences may enhance their usability and enhance their uptake.