Funding Source
National Institute on Drug Abuse – Center for Technology and Behavioral Health Pilot Core
Project Period
May 2022 - May 2023
Principal Investigator
Kammarauche Aneni, MD, MHS
Other Project Staff
Lynn Fiellin, MD (Mentor); Zachary Chase Lipton, PhD (Co-Mentor); Ching-Hua Chen, RN, PhD (Advisor); Youngsun Cho, MD, PhD (Collaborator)
Project Summary
Adolescent substance misuse remains a significant public health problem despite available evidence-based prevention interventions. Adolescents who misuse substances are at increased risk for developing comorbid mental health disorders, engaging in criminal activity, being involved in motor vehicle crashes, and dying by overdose. Early identification of adolescents at risk for substance misuse can stave off these negative consequences. However, most adolescents at risk for substance misuse are not identified early. The use of digital tools such as videogames to identify at risk adolescents offers a potential solution given the wide access to videogames among adolescents and the potential for scalability. Digital biomarkers – “consumer generated physiological and behavioral measures collected through connected digital tools” – can be used to potentially identify adolescents at risk. In-game performance data such as the amount of time to complete a game task, or accuracy of choices in a game task may reflect cognitive processes of executive function such as working memory and inhibitory control which are implicated in the development of substance misuse. As such in-game performance data may be used to identify adolescents at risk for substance misuse. The proposed project aims to identify game-based digital biomarkers of cognitive risk factors implicated in the development of substance misuse among adolescents using available in-game performance data.
In this proposed project, we will use existing participant log data from a videogame developed and evaluated in a randomized controlled trial by the play2PREVENT Lab. We will extract in-game performance metrics that map on to cognitive function and apply machine learning methods to determine which in-game performance metrics – digital biomarkers – more strongly associate with substance misuse. Subsequently, in a pilot sample of adolescents, we will test the association between identified digital biomarkers and executive function.
Findings from this project will inform the development of a game-based tool that could be used in identifying adolescents at risk for substance misuse and related health problems.