Gambling Habit Hacker: A micro-randomised trial of planning interventions delivered via a Just-In-Time Adaptive Intervention for adult gamblers.

Gambling Habit Hacker is a Just-In-Time-Adaptive-Intervention which provides in the moment support for adhering to gambling limits. The aim of this implementation planning intervention is to prevent the deterioration of intention and to support in-the-moment selection of behaviour change strategies and the establishment of robust action and coping plans. Gambling Habit Hacker is informed by the Health Action Process Approach and implementation intentions literature and delivered in accordance with Self-determination Theory. This JITAI employs decision rules specifying that participants who are receptive to treatment (available for treatment) and report low goal strength of intention, low goal self-efficacy, low urge self-efficacy or a high-risk situation (tailoring variables) in Ecological Momentary Assessments (EMAs) sent during three semi-random times a day are delivered action and coping planning activities to implement selected behaviour change strategies. These components are guided primarily by the long-term goal of reducing gambling expenditure (distal outcome), which is posited to be achieved through the short-term goal of adhering to gambling expenditure limits (primary proximal outcome) via increased strength of intention, goal self-efficacy, and urge self-efficacy (secondary proximal outcomes). Trial design A micro-­randomised trial (MRT), a form of sequential factorial design, will be employed to inform the optimisation of this JITAI. In this MRT, each participant will be randomised to an intervention condition or control condition at each decision point across a 28-day period. The MRT will be supplemented with: (1) a within-group follow-up evaluation to explore the longer-term outcomes of the intervention, in which changes to the primary (gambling expenditure) and secondary (gambling symptom severity, gambling frequency, psychological distress, personal wellbeing, situational confidence, and planning propensity) outcome measures from pre-intervention evaluation to post-intervention and six-month follow-up evaluations; and (2) an evaluation of the usability and acceptability of the JITAI using usability and acceptability items in the post-intervention survey of the within-group follow-up evaluation, indices of app usage and engagement, and qualitative telephone interviews. EMA features: All participants will be prompted to complete time-based EMAs (t-EMAs) via push notifications at random times during three pre-specified periods (morning, afternoon, and evening). These in-app t-EMAs will consist of single items assessing strength of intention, goal self-efficacy, urge self-efficacy and presence of a high-risk situation (tailoring variables). The high-risk situations include positive reinforcement high-risk situations (wanting to win back losses), negative reinforcement high-risk situations (temptations to gamble, unpleasant emotions), alcohol or drug consumption, and gambling proximity (whether the person is currently gambling). These items are mostly measured on 5-point Likert scales (from 1-5). The t-EMAs will also include a gambling event record, which will assess gambling expenditure since the last t-EMA completed (e-EMA) (see Primary Outcome). Intervention content: Participants will select from a comprehensive list of cognitive and behavioural strategies to limit or reduce their gambling behaviours. Each option contains a strategy name (e.g., social support) and implementation guidance (e.g., talking to someone about gambling). Each strategy is categorised into a higher-order grouping, which we refer to as a strategy group (n=25). Each strategy group contains between four and six strategies, each with detailed implementation guidance (120 strategies in total). Strategy groups are organised into ten higher-order Behaviour Change Technique (BCT) categories: avoidance, rewards, substitution activities and social support, and categories of BCTs identified in the gambling literature (including maintaining momentum, staying in control while gambling, urge management, financial management, and managing emotions). The action planning component prompts participants to develop personally tailored action plans that respond to immediate threats to adhering to gambling limits. Participants allocated to the intervention condition receive a list of tailored strategy groups (6 to 16 groups) based on the results of their EMA. Participants can select one strategy group, and then the app provides a list of all relevant individual strategies from which participants can select. Upon selection, the app provides a detailed description of methods for implementation (drawn from lived experience research – see consumer involvement section) and offer strategy-specific prompts for the personalisation of the strategy. Once participants are provided with all information and prompted to consider specific details, they are asked to record their action plan. The coping planning component involves the development of a personally tailored coping plan in response to a proximal implementation barrier. Participants are prompted to identify the main barrier right now to action plan implementation by selecting one of seven categories of barriers: thoughts, emotions, motivation, situation, self-belief, financial, and social barriers. Participants are prompted to describe the selected barrier and to detail how that barrier can get in the way of their action plan. The app then prompts participants to identify what they can do right now to overcome the barrier and get back on track with the plan. Once the plan is saved, a commitment and self-efficacy activity follows to facilitate engagement and throughput of the strategy. This involves focusing on character strengths and mental rehearsal. Prompts are provided for consideration of character strengths and imagining themselves implementing their plans. Once completed, the app provides an encouraging message. The total time to complete the control condition is approximately 5 minutes, depending on the level of detail provided by the participant. Intervention options: Each set of t-EMA responses will be evaluated to determine participant eligibility for intervention delivery according to a set of pre-determined decision rules. In total, 17 t-EMA items (measuring strength of intention, goal self-efficacy, urge self-efficacy, and the presence of a high-risk situation) are used to ascertain whether an individual is eligible to receive an intervention. In this app, each EMA item will be linked with multiple strategy groups (e.g., goal self-efficacy item is linked with 12 strategy groups that incorporate strategies to improve goal confidence). Hence, participants who exceed the cut-point for intervention delivery at each t-EMA will be eligible to receive any of the corresponding strategy groups. Participants who do not exceed the cut-point for the relevant t-EMA items will not be eligible for any intervention activities. These participants will receive a short encouraging message and their interaction with the app will end. Importantly, participants who exceed the cut-point for intervention delivery at each t-EMA will be micro-randomised to one of two conditions: (1) control condition (in which participants will select one behaviour change strategy category relevant to their situation and the app closes with an encouraging message) or (2) an intervention condition. The micro-randomisation procedure will involve a 50% chance of receiving the no intervention control condition and a 50% chance of receiving the intervention condition. In the intervention condition, participants will select one behaviour change strategy that is relevant to their current situation. Following selection of a strategy participants select a pre-defined action that they are willing to take in the present moment. Participants are then prompted to form an action plan, identify barriers to plan implementation, and form a coping plan. At the end of each intervention activity, participants are asked to identify a personal strength and to describe how this strength can support implementation intentions. Together these components are expected to take around 5 minutes to complete. The intervention concludes with mental rehearsal aimed at strengthening commitment to action and coping plans. Importantly, Gambling Habit Hacker includes a “provide nothing” option in situations in which the participant is unreceptive, support is not required, or the provision of support may be unsafe, inconvenient, or unethical. Specifically, support will not be offered if the push notification prompting EMA completion is ignored or they press the “snooze” function indicating that are currently unable to complete the EMA (i.e., they are not receptive). Individuals will be allowed two hours to complete an EMA to preserve the momentary nature of the intervention while accommodating the potential for possible unavailability (e.g., driving, working) of the participant at the initial notification time. Notably, the micro-randomised trial will be conducted over a 28-day period. During this time, intervention activities will only be available in response to the push notification (“push” intervention). These notifications will not be delivered during the six-month follow-up period, but the intervention content will be available to participants “on-demand” during this time (tailored “pull” intervention). During this period of time, participants will be able to access the same tailored intervention activities via a participant-initiated t-EMA.