Boston VA tests AI for Veteran care

The VA Boston Healthcare System recently hosted an enterprise-wide artificial intelligence innovation workshop exploring how generative AI tools might help clinicians diagnose and treat post-traumatic stress disorder more efficiently. The pilot project, led by researchers at VA's National Center for PTSD, tested whether AI could assist providers in analyzing the complex clinical information gathered during PTSD assessments, which currently require up to an hour to administer and additional time to interpret. The goal is not to replace clinical judgment but to help providers reach more veterans by streamlining the diagnostic process.

PTSD diagnosis within the VA system relies on structured, psychometrically valid assessment tools like the Clinician-Administered PTSD Scale for DSM-5, known as CAPS-5. These assessments generate large amounts of clinical data that providers must synthesize while applying their professional expertise and judgment. With thousands of veterans seeking PTSD care each year, the time-intensive nature of these evaluations can create bottlenecks in access to treatment. The VA is now investigating whether AI can organize and present this information more efficiently, allowing clinicians to spend more time listening to veterans and making informed care decisions rather than processing raw data.

This development affects any veteran seeking mental health care through the VA, particularly those waiting for PTSD evaluations or treatment. While the article does not specify disability rating tiers or geographic limitations, the innovation workshop at VA Boston is designed to develop solutions that could eventually be scaled across the entire VA healthcare system. Veterans currently enrolled in VA healthcare who are seeking mental health services stand to benefit if these AI tools prove effective and are implemented more broadly.

Early testing of the PTSD diagnostic tool showed it performed at a level comparable to a clinician, according to Dr. Amar Mandavia, a clinical psychologist and informatician with the National Center for PTSD. However, the team also discovered important limitations: adding more training data made the tool too rigid, and some of the project's larger ambitions exceeded what current AI models and resources could support. The workshop also explored other applications including operating room scheduling, veteran transportation, research support, and various clinical and administrative workflows, with each project beginning with challenges identified by frontline VA employees.

For disabled veterans, this matters because access to timely mental health care directly impacts disability claims, treatment outcomes, and quality of life. PTSD is one of the most common service-connected disabilities, and delays in diagnosis can mean delays in appropriate treatment and compensation. If AI tools can help clinicians work more efficiently without sacrificing accuracy, more veterans could receive evaluations and begin treatment sooner. The VA's emphasis on keeping human judgment at the center of decision-making also addresses legitimate concerns about whether technology might depersonalize care for conditions that require nuanced clinical expertise.

Veterans currently seeking PTSD evaluations or mental health services through the VA should continue working with their providers through existing channels, as these AI tools remain in the testing phase. Those experiencing delays in accessing mental health care may want to discuss their situation with a Veterans Service Officer, who can help navigate the system and ensure claims are properly documented. As the VA continues developing and testing these technologies, staying informed about changes to assessment processes and asking providers about new tools being implemented at your local VA medical center can help you understand how care delivery may evolve.