The Office of the National Coordinator for Health IT used SAMHSA data to assess health IT adoption in behavioral health settings and found wide variability in electronic health record use across clinical and administrative tasks. Despite rising demand for mental health and substance use services, many facilities lag in exchanging behavioral health data, creating persistent interoperability gaps that limit care coordination and reporting.
Those data exchange shortfalls have direct implications for AI in behavioral healthcare: fragmented records and uneven EHR adoption constrain training datasets, hinder real-time decision support and predictive analytics, and increase risk of algorithmic bias. Confidentiality rules and inconsistent standards further complicate sharing. ONC’s findings underscore the need for standardized APIs, robust consent frameworks and targeted policy and technical work to enable secure data flows that would support validated AI tools, population health analytics and improved clinical workflows in behavioral health.
