Background Intervention development is a critical first step when conducting trials or observational studies. Research suggests that involving multiple stakeholders in this process increases the likelihood of developing interventions that are acceptable, engaging, feasible and effective. However, there is little guidance for researchers on the best ways to involve multiple stakeholders in a meaningful way. The aim of this Study Within A Trial (SWAT) is to identify the most suitable approach to involve patients and healthcare professionals in a consensus process to inform the development of the Improve Diabetes Eye-screening Attendance (IDEA’s) intervention.

Methods This is a qualitative study. Three meetings were held to establish consensus on the content and delivery of the intervention. Meeting included 1 patients only, meeting 2 included a combination of patients and healthcare professionals and meeting 3 included healthcare professionals only. Stakeholders were asked to agree on intervention components which target patients and general practices. Each meeting was audio recorded and field notes were taken. After the meeting, semi-structured telephone interviews were carried to explore stakeholders’ experiences of taking part. Data were transcribed verbatim and managed using NVivo V12 software. Thematic analysis was performed to identify themes relating to members’ experiences of taking part.

Results All three meetings put forward feasible ideas which were incorporated into the final intervention. Each meeting had ideas which were not put forward by the other groups but were incorporated into the final intervention (Meeting 1=6 ideas, Meeting 2=2 ideas and Meeting 3=5 ideas). Thematic analysis suggests that stakeholders in meetings 1 and 3 felt comfortable expressing their opinions and grateful that they were given the opportunity to be heard. Stakeholders in meeting 2 felt they had to hold back on their opinions as they were aware that the other stakeholder group was in the room. They also felt that their contributions were undervalued by the other stakeholder group.

Conclusion Involving patients and healthcare professionals together in a consensus process is not a suitable approach to involvement. This study will guide researchers on the most suitable approach to involve patients and healthcare professionals in a consensus process and will contribute to the evolving literature on the potential impact of involving multiple stakeholders in the intervention development process.