Mental Health Mobile Apps for Adults

Over 10,000 mental health mobile apps are commercially available and counting. Only a small number of apps are regularly downloaded and used, which is in stark contrast with the surge of depression and anxiety apps on the marketplace. Most commercial apps were released without any supportive data. It is costly to develop a brand-new app, which requires a professional team to maintain. The reach and impact are questionable. Some apps include rich contents, yet the public or patients may find them not useful. The preferences of mobile apps may vary in different populations. A list of the high-quality mobile apps will be formed, then conduct a pilot randomized controlled trial (RCT) to evaluate whether a brief advice to provide these apps is feasible for adult populations with mild to moderate depressive symptoms. The assumption is that digital-based self-help interventions are effective methods to improve wellbeing and promote mental health, and their impact could be extended at the population level with a minimal cost and time if proved to be effective. In this feasibility trial, the research aim was to explore trial design, subject acceptability of the interventions and preliminary effectiveness of the interventions, and to provide data to estimate the parameters required to design a definitive RCT.

A parallel-group pilot RCT with an equal randomization of 1:1, (1) Intervention group: brief advice of a list of mobile apps (LMA) and (2) Waitlist control (WL) group. A total of 234 participants (117 in each group) who have mild to moderate depression or anxiety symptoms will be recruited. Participants will be approached through mass email, public posters or active approach in community centres. The randomisation will be conducted after the participants give consent and complete the baseline assessments (T1) (via Qualtrics). The individuals will be randomised into intervention and control groups. The randomized list will be generated using an Excel spreadsheet by a statistician who is not involved in any part of the study. The recruitment staff will be blinded in the allocation process. The randomisation number will not be decoded to the participants until they get assigned a group. Considering the nature of the intervention, participants and research staff cannot be blinded in the process. As the recruitment will be mainly conducted online and through telephone, contamination between groups is expected to be minimal. Potential contamination will be checked at the T1 assessment by asking whether any participants they know have attended the other group. In the final assessment, the participants in the control group will be asked whether they have used any of the recommended apps in the past 3 months.

A 2-hour instruction session of the mobile apps will be delivered to all the participants via Zoom or other online platforms. Participants will be introduced to the concept of self-care and presented with a list of four high-quality and suitable mobile apps along with their main functions. Participants will be instructed to use at least one of the apps for 10 minutes daily over a period of 30 days. They can choose one app or a combination of two or three apps. Participants will create their own user accounts for the chosen app(s), with a reminder to prioritize privacy. Two reminders will be sent during the second and third weeks, emphasizing app usage, the concept of self-care, and providing tips for using the apps. If participants encounter any difficulties accessing the apps, the project coordinator will discuss the issues and collaborate with them to find solutions. Participants will complete the process evaluation (intervention group only) immediately after the intervention (T2), complete the process evaluation (intervention group only) and outcome assessment at 3-month (T3). A fidelity checks on adherence and competence will be conducted.

The feasibility outcomes, including the recruitment rate and duration, retention rate, and adverse events, will be reported descriptively and narratively. For the clinical endpoints, descriptive statistics, mean (standard deviation) for continuous outcomes, will be reported. After adjusting the baseline covariates, mean differences in the outcomes will be compared between the two groups. Multiple imputation approach will be used to deal with missing data (mice package in R). Three sensitivity analyses will be carried out to assess the effect of three factors on the outcomes (retention, choice of mobile apps, effective features). For the in-depth telephone interview, content analysis method will be used following the framework method analysis procedure (e.g., coding). It is expected that the intervention group show more significant improvements in the measured outcomes than the control group, indicating preliminary evidence of effectiveness.