AI-Enhanced App-based Intervention for Adolescent E-cigarette Cessation

Project Summary Abstract Electronic cigarette (e-cigarette) use among adolescents is an emerging public health problem. In 2020, 19.5% of high school students and 4.7% of middle school students used e-cigarettes, even though e-cigarette use is known to be harmful to adolescent health, resulting in nicotine-related symptoms of addiction and

various other health issues (e.g., depressive symptoms), and is associated with use of other addictive substances. In spite of the urgent need to address this issue, there is a lack of effective interventions for adolescent e-cigarette cessation. Thus, effective and sustainable interventions to address adolescent e-

cigarette use are urgently needed. This study is aimed at creating a scalable e-cigarette use cessation intervention using a smartphone application (app), which can be delivered to a large pool of adolescents. This study will test and launch a scalable smartphone AI-enhanced app-based intervention delivery that can be

feasibly implemented in schools, primary care clinics, and community settings for adolescent e-cigarette cessation. An artificial intelligence (AI)-enhanced smartphone app-based intervention is a cost-effective strategy that facilitates personalized messages for behavioral problems. Using AI-assisted user profiling

techniques based on participants’ e-cigarette use status, patterns, other substance use status, readiness to quit e-cigarettes, and beliefs about e-cigarettes, this intervention will send personalized content with mindfulness-based training modules. Timely messages will be sent in response to users’ urges to use e-

cigarettes based on self-report in real time through the smartphone app. In addition, AI will be able to answer app users’ questions via an automated chatbot. We will pilot test the intervention for summative evaluation using a quasi-randomized controlled study using a delayed intervention design. In Phase 1, we will refine a

smartphone app-based intervention based on feedback obtained from participant interviews on usability (functions, interface, and content), a usability survey, and a task performance test. In Phase 2, we will pilot test the intervention using a quasi-randomized controlled study design. We will examine the usability via a usability

survey, and participants’ engagement on the app use via frequency of AI-app use (each feature), data entered in the app, and minutes of app use per day. In addition, we will examine the e-cigarette cessation related outcome changes of the app-based intervention by evaluating e-cigarette use frequencies compared to the

control group (3 month delayed intervention) in terms of quit rates at post-intervention, 30-days and 3-months post-intervention; number of days e-cigarette used in the past 30 days; number of quit attempts; intention to quit; readiness to quit; nicotine addiction; and beliefs and perceptions about e-cigarettes. We will examine the

feasibility of the intervention based on user experiences of functions, interface, content. Over the long term, findings from this study will contribute useful information for those testing the effectiveness of smartphone app- based interventions for adolescent e-cigarette use cessation on a large scale.