A systematic review of chatbot-assisted interventions for substance use

Objectives: This study systematically reviewed research on the utilization of chatbot-related technologies for the prevention, assessment, and treatment of various substance uses, including alcohol, nicotine, and other drugs. Methods: Following PRISMA guidelines, 28 articles were selected for final analysis from an initial screening of 998 references. Data were coded for multiple components, including study characteristics, intervention types, intervention contents, sample characteristics, substance use details, measurement tools, and main findings, particularly emphasizing the effectiveness of chatbot-assisted interventions on substance use and the facilitators and barriers affecting program effectiveness. Results: Half of the studies specifically targeted smoking. Furthermore, over 85% of interventions were designed to treat substance use, with 7.14% focusing on prevention and 3.57% on assessment. Perceptions of effectiveness in quitting substance use varied, ranging from 25% to 50%, while for reduced substance use, percentages ranged from 66.67% to 83.33%. Among the studies assessing statistical effectiveness (46.43%), all experimental studies, including quasi-experiments, demonstrated significant and valid effects. Notably, 30% of studies emphasized personalization and providing relevant tips or information as key facilitators. Conclusion: This study offers valuable insights into the development and validation of chatbot-assisted interventions, thereby establishing a robust foundation for their efficacy.