30 Development and evaluation of the RECAP trial mobile application and web portal: preliminary testing on healthy volunteers

<h3>Background</h3> RECAP trial is the first feasibility trial to use Artificial Intelligence (AI) to automate exercise prescription as part of a home-based programme in the NHS. <h3>Aim</h3> We describe the design and development of a mobile app and web-portal, and the evaluation of the prototypes via testing on healthy staff volunteers at the James Cook University Hospital. <h3>Methods</h3> A prototype of the app was developed using the Flutter Framework for the iOS and Android environments. Thirteen cardiac research and rehabilitation staff were enrolled on to the rehabilitation programme for two weeks. Participants’ and physical activity data were collected using accelerometers for one week and exercise goals were prescribed using AI. At the end of the programme, participants completed the System Usability Scale (SUS) questionnaire and a questionnaire asking about their experience of the app. Ten participants who took part in the mobile app testing stage were given the opportunity to use the web portal, and usability testing was conducted using a ‘thinking-aloud’ interview approach. <h3>Results</h3> Thirteen participants (mean age 50, 12 females) took part in the study. Mean step count per day of participants increased from 6295 in week one to 6571 in week two (table 1). The mean SUS score was 78.6, corresponding to a grade of B+. In response to the questionnaires, participants found the goals set by AI algorithm were appropriate and challenged them to do more. Mobile app analytical data demonstrated that each participant had an average daily engagement time of 1 minute and 25 seconds. Qualitative interviews on the usability of the web portal found that it was generally easy to use but some areas of navigation were not well defined and needed improvement. <h3>Conclusion</h3> The mobile app and web portal were found to be usable and were further improved before the clinical trial.