Pilot trial of Mello: Investigating the effect of a targeted, personalised smartphone app on repetitive negative thinking in young people with depression and anxiety

Participants in the intervention arm will be given access to a smartphone app called Mello for a total of 6 weeks, which will be downloaded onto their own iPhone or Android device. They will be free to use Mello as much or as little as they want during this time. Mello is a world first smartphone intervention targeting repetitive negative thinking (RNT); a key transdiagnostic process underpinning depression and anxiety. RNT involves repeatedly going over negative thoughts in the mind in a way that is passive and/or difficult to control, resulting in reinforcement of negative mood states. As RNT is a mechanism which has been found to drive depression and anxiety, interventions that effectively reduce RNT may have a flow on effect for improving depression and anxiety outcomes. Mello offers a means of disrupting RNT as it unfolds in the moment by delivering personalised interventions at key moments in daily life via a smartphone app. Delivering momentary interventions in daily life is referred to as ecological momentary intervention (EMI). Via an interactive chat interface, users are prompted three times a day to complete a short four item questionnaire (‘check in’) via the app to gather contextual information about current levels of RNT, mood, surrounding environment and activities. Each item is rated on a 5 point Likert-type scale from 1 (not at all) to 5 (very much). Collecting repeated data of momentary experiences in daily life in this way is referred to as ecological momentary assessment. Responses to the questionnaire are used to tailor suggestions for the user to complete one of 12 possible therapeutic activities designed to disengage young people from RNT in the moment. In this way, Mello offers personalised intervention strategies matched to the level of stuck thinking (high or low), mood (positive or negative), context (at home or away from home), and activity (active or passive). Time of day (day or night) is also collected from the app and used to tailor the therapy activity recommendation. These rules underpinning which therapeutic activities are recommended to young people were determined based on reviews of the literature on RNT mechanisms, and input from clinicians and young people with lived experience. In addition to prompting therapy activities via the check-ins, participants can access all therapy activities on demand at any time via another section of the app contained within a profile, where they can also save their favourite therapy activities. Therapy activities were derived from a meta-analysis of interventions targeting RNT in young people with depression and anxiety (Bell et al., in prep), as well as workshops with young people and clinicians outside the research group. A team of digital technology developers, researchers and clinicians within the research group then worked together to translate the evidence based intervention strategies identified from this meta-analysis into a digital format, following the requirements that these activities were: 1) engaging, 2) brief (5-10 minutes), and 3) likely to disrupt RNT in the moment. Further user testing with young people and clinicians then informed development and refinement of the therapy activities with a primary aim to ensure excellent user experience. These interactive, multimedia based activities primarily target disrupting the process of RNT by encouraging mindful detachment from negative thoughts through a number of cognitive behavioural strategies, including mindfulness, defusion, worry time, problem solving and attention training. Each activity takes approximately 5 minutes and involves guided audio or text options, with additional multimedia elements including interactive text entry, images and animations. Following completion of a therapy activity, the participant is returned to the chat interface which collects follow up information about the level of RNT and ratings of how helpful the activity was. Contextual information collected within the check-in questionnaire feeds into a profile section of the app which provides insights into the relationship between RNT and contextual factors (mood, activity, location). This information is displayed as variables reported when stuck thinking is reported to be low (i.e. scores 1 or 2 on the check in scale) or high (i.e. scores 3 to 5 on the check in scale). The information on this page is only revealed once at least 7 check-ins are completed (33% of total number possible check-ins within one week) and at least one week of app use is completed, to ensure the feedback is representative of a spectrum of experiences. This component of the app targets improved self-awareness via self-monitoring and feedback, a key behavioural change strategy with evidence supporting clinical benefits for reducing RNT and improving mood conditions. In addition to using Mello, engagement will be supported by weekly calls which will last approximately 15 minutes to participants from research assistants who check in on and encourage usage, and troubleshoot any problems. Whilst engagement with the app will be included as an outcome measure, this data will not be monitored during the course of the trial and addressed with participants. Changes to the app during the course of the trial may be necessary to maintain optimum performance and app stability. In line with Mohr et al (2015), core active components of the intervention will not change during the trial. These include: 1) Notification and check in process, including the frequency of prompts and content of check ins, as well as the ability to check in on demand 2) Therapy activities and logic, including all content of the therapies, the logic linking check in contextual variables to activity recommendations 3) Ability to access therapy activities on demand via the home screen button, save favourites and schedule reminders for activities 4) About you screen displaying links between level of stuck thinking and contextual variables (i.e. check in data) Features of the technology which directly affect usability (e.g. bug fixes, performance issues) will be addressed continuously during the trial but all core features will remain constant. Mohr, D. C., Schueller, S. M., Riley, W. T., Brown, C. H., Cuijpers, P., Duan, N., ... & Cheung, K. (2015). Trials of intervention principles: evaluation methods for evolving behavioral intervention technologies. Journal of medical Internet research, 17(7), e166.