Usage and Differential Effectiveness of a Mobile App to reduce substance use, gambling and digital media use in vocational school students: Secondary Analysis of the Intervention Arm of a Randomized Controlled Trial (Preprint)

Abstract

During adolescence, substance use and digital media exposure usually peak and can become major health risks. Prevention activities are mainly implemented in the regular school setting and youth outside this system are not reached. A mobile app (“Meine Zeit ohne”-app) has been developed specifically for vocational students and encourages participants to voluntarily reduce or abstain from a self-chosen addictive behaviour including use of a substance, gambling, or a media-related habit such as gaming or social media use for 2 weeks. Results from a randomized study indicate a significant impact on health-promoting behavior change after using the app. The present secondary study focuses on the intervention arm of the previous study focusing on acceptance and differential intervention effectiveness based on app use and personal characteristics.

The aims of this study were (1) to examine characteristics of participants who used the app, (2) to explore the effectiveness of the mobile intervention depending on how the app was used and depending on individual participants’ characteristics, and (3) to study how variations in app use were related to participants’ baseline characteristics.

App usage was analyzed based on log data from study participants in the intervention group and included frequency of app use (in days), selection of specific challenge and personal relevance (i.e., user was above a predefined risk score for a certain addictive behavior) of challenge selection (“congruent use”: e.g., smoker selected challenge related to reducing or quitting smoking). Dichotomous outcomes (change vs no change) referred to past month substance use, gambling, media-related behaviors. The relationship between the usage metrics, personal relevance and change in outcome variables was analyzed using binary multilevel mixed-effects logistic regression models.

The intervention group consisted of n = 2367 vocational students; n = 1458 (mean age 19.22 years, 55.3% male) of them provided full data, and n= 894 (61.3%) started a challenge and could be included in the analysis (18.7 years, 43.9% female). N = 466 (52.1%) students were considered frequent app users with more than four days of active use over the 2-weeks period. The challenge area most often chosen in the analyzed sample was related to social media use (n= 332, 37.1%). N = 407 students (45.5%) selected a challenge in a behavioral domain of personal relevance. Effects of app use on outcomes was higher when the area of individual challenge choice was equal to the area of behavior change, challenge choice was related to a behavior of personal relevance, and the individual risk of engaging in different addictive behaviors was high.

The domain-specific effectiveness of the program was confirmed with no spillover between behavioural domains. Effectiveness appeared to be dependent on app use and users’ characteristics.

DRKS00023788

RR2-https://trialsjournal.biomedcentral.com/articles/10.1186/s13063-022-06231-x