NIHCM’s September webinar featured an in-depth discussion with leaders on mental health trends and the care delivery landscape. In this exclusive interview, John Torous, MD, MBI, Director of the Digital Psychiatry Division at Beth Israel Deaconess Medical Center, shares effective ways to use digital technology in care, focusing on how AI can support clinicians and expand access, particularly in underserved communities, while also addressing critical safety concerns surrounding AI chatbots for youth.
Q: Where do you see the greatest potential for AI to improve access to and delivery of mental health care? Dr. Torous posits that AI's primary role is to augment, rather than replace, human clinicians, thereby enabling the delivery of enhanced care to a larger population. He envisions AI as a supportive tool for patients between clinical appointments, offering insights into subtle changes in symptoms and behavior, and facilitating connections to appropriate resources. Research conducted by his division on hybrid care models, digital phenotyping, and the role of Digital Navigators consistently indicates that technological interventions are most effective when seamlessly integrated with human support. The core opportunity lies in transforming mental health care into a more continuous, personalized, and broadly accessible service, all while preserving the crucial human element in patient-provider relationships. Furthermore, he suggests that this integration could pave the way for actual preventative mental health strategies. Q: How might AI expand mental health care in rural and underserved communities, and what barriers may limit adoption? Dr. Torous suggests that AI holds significant promise for extending the reach of scarce mental health resources, particularly in rural and underserved areas. It could achieve this by empowering frontline healthcare workers, offering crucial educational content, and assisting individuals in navigating the often-complex landscape of available services. This application would be especially beneficial in regions where mental health specialists are few and far between. However, he cautions that simply providing access to technology does not automatically guarantee access to quality care. Significant barriers to adoption include disparities in digital literacy, inconsistent internet connectivity, language barriers, issues of cultural relevance, privacy concerns, and a fundamental lack of trust in new technologies. His team's work with Digital Navigators underscores the vital importance of human-centered support in helping individuals effectively and equitably utilize digital health tools. Q: Can you share any key insights or takeaways on the safety of AI chatbots for youth mental health, and what stakeholders looking to ensure safe AI use and implementation need to know? Dr. Torous highlights a growing trend where young individuals are independently seeking mental health support from AI chatbots, often outside the traditional healthcare system. He emphasizes a critical distinction: AI tools provide support, but this is not synonymous with professional therapy. While these AI tools can offer certain benefits, there are rare but serious risks, including the potential for inaccurate advice, the reinforcement of harmful beliefs, or a failure to adequately recognize and respond to mental health crises. A key insight from their research is that evaluating the safety of these tools requires more than isolated test questions. Instead, safety must be rigorously assessed through analysis of real-world conversations over extended periods, with active input from youth perspectives. He stresses the necessity of independent oversight and the establishment of clear, accessible pathways to human professional support to ensure responsible and safe AI implementation in youth mental health.
Q: Where do you see the greatest potential for AI to improve access to and delivery of mental health care?
Dr. Torous posits that AI's primary role is to augment, rather than replace, human clinicians, thereby enabling the delivery of enhanced care to a larger population. He envisions AI as a supportive tool for patients between clinical appointments, offering insights into subtle changes in symptoms and behavior, and facilitating connections to appropriate resources. Research conducted by his division on hybrid care models, digital phenotyping, and the role of Digital Navigators consistently indicates that technological interventions are most effective when seamlessly integrated with human support. The core opportunity lies in transforming mental health care into a more continuous, personalized, and broadly accessible service, all while preserving the crucial human element in patient-provider relationships. Furthermore, he suggests that this integration could pave the way for actual preventative mental health strategies.
Q: How might AI expand mental health care in rural and underserved communities, and what barriers may limit adoption?
Dr. Torous suggests that AI holds significant promise for extending the reach of scarce mental health resources, particularly in rural and underserved areas. It could achieve this by empowering frontline healthcare workers, offering crucial educational content, and assisting individuals in navigating the often-complex landscape of available services. This application would be especially beneficial in regions where mental health specialists are few and far between. However, he cautions that simply providing access to technology does not automatically guarantee access to quality care. Significant barriers to adoption include disparities in digital literacy, inconsistent internet connectivity, language barriers, issues of cultural relevance, privacy concerns, and a fundamental lack of trust in new technologies. His team's work with Digital Navigators underscores the vital importance of human-centered support in helping individuals effectively and equitably utilize digital health tools.
Q: Can you share any key insights or takeaways on the safety of AI chatbots for youth mental health, and what stakeholders looking to ensure safe AI use and implementation need to know?
Dr. Torous highlights a growing trend where young individuals are independently seeking mental health support from AI chatbots, often outside the traditional healthcare system. He emphasizes a critical distinction: AI tools provide support, but this is not synonymous with professional therapy. While these AI tools can offer certain benefits, there are rare but serious risks, including the potential for inaccurate advice, the reinforcement of harmful beliefs, or a failure to adequately recognize and respond to mental health crises. A key insight from their research is that evaluating the safety of these tools requires more than isolated test questions. Instead, safety must be rigorously assessed through analysis of real-world conversations over extended periods, with active input from youth perspectives. He stresses the necessity of independent oversight and the establishment of clear, accessible pathways to human professional support to ensure responsible and safe AI implementation in youth mental health.
