Detour: a Smartphone Game to Help Youth Quit Smoking

The aim of the present two-armed randomized controlled trial is to test the efficacy of Detour, a smartphone game to help youth (16-25 years) quit smoking. It is designed to be an accessible, multi-component intervention providing in-the-moment distractions from cravings. First, Detour targets stimuli devaluation using a gamified go/no-go training in an endless runner game format. Second, the researchers aimed to reduce delay discounting in two ways; 1) players can replace neutral images in the go/no-go training with positive images from their own lives related to their future self; 2) players can set and monitor personal goals via the game, to promote future-self priming. Third, participants join a social shell comprising Instagram and TikTok content to provide psychoeducation about quitting smoking and to interact with fellow players who quit simultaneously. Finally, the timing and content of in-game messaging is personalized to each participant based on information provided in an ecological momentary assessment component and a stop plan filled out during the week before the quit attempt.

Detour will be tested against a self-help digital brochure, "Jouw eigen plan om te stoppen met roken" by the Trimbos Institute, the Netherlands. The brochure provides guidelines to independently design a stop-plan including questions about seeking support from others, strategies to fight cravings, and setting rewards for milestones after quitting. Additionally, the brochure has psychoeducational information that is valuable when planning to quit smoking.

For the data collection procedure, the researchers will first broadly distribute a short screening questionnaire via educational institutes and social media to assess study inclusion criteria. Eligible participants will be contacted by phone/text message/email to explain the purpose of the study further and invited to fill out the pre-test measurement if they would like to participate. The researchers will recruit nationwide and strive for a diverse sample in terms of gender, demographic location and educational levels. Participants will provide informed consent at both the screening and pre-test phases.

After the pre-test, blocks of 6 participants are randomly assigned to either the "Detour" game group or the brochure group to begin the intervention. The intervention period lasts for five weeks. During week 0 or the preparatory week, participants can access Detour or the brochure but have not yet quit. On day 1 of week 1, known as quit day, participants will quit smoking and try to remain smoke-free for the next four weeks (weeks 1-4).

The target sample size is N = 604, with 10% drop-outs taken into account. Recruitment will stop once 604 participants have received an intervention or on August 1, 2025.

The aim is to achieve 25% abstinence rates at 6-month follow-up in the game group. The researchers aim to prove a small effect size (odds ratio 1.88) with 90% power. A power analysis (logistic regression analysis, one-tailed alpha .05, and a fifty-fifty distribution of participants in both groups) showed that they will need N = 604 participants, taking into account a 10% loss.

To assess the immediate effectiveness of the game for smoking cessation in youth, logistic regression analyses will be performed with intervention group (game or brochure), as predictor and 14-day point prevalence of smoking abstinence (yes/no) at 6-month follow-up as the outcome. Age, Gender, socioeconomic status (SES), and baseline weekly smoking will be added to the model as controlled variables.

To assess the longer-term effectiveness of the game for smoking cessation in youth, logistic regression analyses will be performed with intervention group (game or brochure), as predictor and 14-day point prevalence of smoking abstinence (yes/no) at post-test as the outcome. Age, Gender, socioeconomic status (SES), and baseline weekly smoking will be added to the model as controlled variables.

To assess the effect of the game on weekly smoking, multilevel regression analysis will be performed with intervention group (game or brochure), time (pre-test, post-test, 3-month follow-up and 6-month follow-up), and their interaction as predictors, and weekly smoking as the outcome. The following models will be tested. Age, Gender, socioeconomic status (SES), and baseline weekly smoking will be added to the model as controlled variables.

• Model0 ← WeeklySmoking ~ (1 | ID); check intra-class correlations
• Model1 ← WeeklySmoking ~ TreatmentGroup + Time + Gender + Age + SES + Baseline weekly smoking + (1 | ID)
• Model2 ← WeeklySmoking~ TreatmentGroup + Time + Gender + Age + SES + Baseline weekly smoking + (1 + time | ID); compare model 1 and model 2 to see if random slopes give better model fit.
• Model3 ← WeeklySmoking~ TreatmentGroup + Time + TreatmentGroup:Time + Gender
• Age + SES + Baseline weekly smoking + (1 + time | ID); Calculate simple slopes for both cases of treatment group.

For the analyses of the primary and secondary outcomes, a one-tailed p < .05 criteria will be used. For remaining analyses the standard p < .05 criteria will be used.

Furthermore, the intention to treat approach will be used, where it is assumed that people who missed post-test or either follow-up are non-abstinent.

Participants who only responded to the pretest but neither post-test, 3-month follow-up, nor 6-month follow-up will be considered as relapsers in analyses with abstinence as the outcome. In models with weekly smoking as the outcome, the majority of weekly smoking values would need to be imputed, so instead participants who only responded to the pre-test measure will be excluded.

The present pre-registration only pertains to the efficacy of the intervention. A follow-up registration on the in-depth analysis of the intervention effects and sub-group analyses to test who the intervention worked best for will be made. Please refer to the project page for more information (https://osf.io/k784a/).