Machine Learning in Health Promotion and Behavior Change: A Scoping Review

Background: Despite health behavior change interventions targeting modifiable lifestyle factors underlying chronic disease, dropouts and non-adherence by individuals have remained high. The rapid development of machine learning (ML) in recent years, along with its ability to provide an easily accessible personalized experience to users, holds great potential for success in health promotion and behavior change interventions. .

Objective: The objective of this article is to provide an overview of existing research on the applications of ML and to exploit their potential in health promotion and behavior change interventions.

Methods : A scoping review was performed based on Arksey and O’Malley’s 5-step framework and Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Scoping Reviews (PRISMA-ScR) guidelines. A total of 9 databases (the Cochrane Library, CINAHL, Embase, Ovid, ProQuest, PsycInfo, PubMed, Scopus and Web of Science) were consulted from their creation until February 2021, with no limit on dates and types of publications. Studies were included in the review if they incorporated ML into health promotion or behavior change interventions, studied at least one group of participants, and were published in English. Publication information (author, year, purpose and results), health promotion field, user data analyzed, type of ML used, challenges encountered and future research were extracted from each study .

Results: A total of 29 articles were included in this review. Three themes were generated, which are: (1) enablers, ie the adoption of information technology to optimize system functioning; (2) challenges, which include the various obstacles and limitations presented in the articles; and (3) future directions, which explore prospective strategies for health promotion through ML.

Conclusion : The challenges were not only the time and resource consuming nature of ML-based applications, but also the burden on users for data entry and the degree of customization. Future work could consider designs that accordingly mitigate these challenges in areas that receive limited attention, such as smoking and mental health.

Comments are closed.