AppReminders - a pilot feasibility randomized controlled trial of a memory aid app for people with acquired brain injury.

Only 29 people were randomised instead of the planned 50, and just 19 finished with usable memory data. The trial was not set up to compare the two apps, so the memory differences it reports are exploratory and could easily be chance. Baseline memory scores were already very high (86% on the memory log), close to the ceiling, which leaves almost no room for any app to show an improvement. The scoring of the memory log changed partway through: an extra point for remembering without being prompted was added after four people had already enrolled, so 16 people were scored on the newer version and the rest were not. People who stayed in the study were probably more motivated and functioning better than average. Most of the dropouts happened before the training session, so the people who reached the app were self-selected. Only 63 of the 150 eligible service users were referred by the services, which leaves open the possibility that staff put forward the people they thought would cope. Participants were not blinded to the purpose of the trial and were told which apps were being compared, so the more favourable ratings of ApplTree may reflect knowing the researchers had built it. The text-message task may have measured how engaged people were with the study rather than their memory, and some follow-up memory logs were scored by researchers who knew which group the person was in. The trial ran across the Covid-19 pandemic, with a nine-month pause and a move to phone and video sessions, which the authors say makes the results less generalisable. Follow-up was only three weeks after six weeks of app use. Memory aids are meant to be used long term, so nothing here speaks to whether people keep using an app over months or years. The people who coded the app feedback were the researchers who designed and delivered the intervention; a second researcher checked 20% of it.

Funded by a Chief Scientist Office Translational Clinical Studies Research Committee award (a Scottish Government health research funder) and pre-registered on ClinicalTrials.gov before recruitment. The supplied text carries no competing-interests declaration. The app being tested, ApplTree, was developed by the research team, and the authors state that the feedback was coded by the researchers who designed and delivered the intervention.

Do not read the 13% rise in the ApplTree group's memory log scores as proof that a better-designed app improves memory. The trial was not powered to compare the two apps, the combined memory measure showed no change, the difference disappeared once remembering without a prompt was scored, and participants were not blinded to the fact that the researchers had built ApplTree.