Indianapolis, IN — A new partnership led by the Regenstrief Institute and the U.S. Department of Veterans Affairs is building a learning health system designed to improve care for older veterans by using real-time data to guide clinical decisions. The effort, developed with the VA’s Center for Health Information and Communication in Indianapolis and Indiana University School of Medicine, is targeting chronic conditions common among veterans over 65, including dementia, mobility limitations, chronic pain, and social isolation.
The initiative expands on a VA Advanced Fellowship program launched in 2021 and now incorporates artificial intelligence and patient-reported outcomes to create rapid feedback loops between the clinic and the front lines. Early pilots are underway at VA facilities in Indiana and the broader region, with plans to scale nationally by 2027. Initial funding includes a $2.5 million VA Health Services Research & Development grant and support from the National Institute on Aging.
More than 70% of veterans enrolled in VA care are 65 or older, according to VA demographics, intensifying pressure to modernize how geriatric care is delivered across the system. “This isn’t just about data—it’s about dignity and empowerment for those who’ve served,” said Regenstrief CEO Dr. Peter Embí. “By building a learning health system, we’re ensuring care learns and adapts in real time.”
How the system works
At its core, a learning health system continually aggregates and analyzes clinical information—pulling from electronic health records, caregiver input, and veterans’ own reports—then pushes evidence-based prompts back to clinicians and care teams. In the VA, this work augments longstanding platforms like VistA and supports the agency’s ongoing EHR modernization, while adding tools that can flag patients at risk of falls or hospitalization and personalize interventions such as telehealth visits, home-based physical therapy, or social supports.
Regenstrief researchers say the approach aligns with the “triple aim” of improving patient experience, population health, and cost of care. It also builds on VA programs such as Gerofit, which has adapted exercise and mobility support for remote delivery to reach rural veterans.
Early results and equity focus
According to institute data, pilot sites have recorded a 15% to 20% drop in hospital readmissions for chronic pain management, with notable gains for Black veterans—a group that previous VA studies found faced higher rates of fragmented care. In one case example described by project leaders, a 78-year-old Vietnam veteran with musculoskeletal pain reduced his opioid use by about 30% after receiving an AI-guided, personalized exercise plan.
VA leaders say the work builds capacity inside the system. “The partnership is pivotal in our Advanced Fellowship Program,” said Dr. David Atkins, who oversees VA health services research. “It orients fellows to learning health principles and prepares them to transform care for an aging population.”
Timelines, funding, and safeguards
The project’s next phase includes expanding pilots in 2026, adding dementia prediction tools and more robust caregiver integration, before a planned VA-wide rollout by 2027. Beyond VA HSR&D and NIA backing, the program is coordinating with local VA medical centers, community partners, and veteran service organizations to ensure the data’s lessons translate to policy and practice.
Leaders acknowledge ongoing challenges: integrating legacy systems, ensuring interoperability under the 21st Century Cures Act, and safeguarding data privacy under HIPAA. Project teams say they are building modular tools that can be adopted site by site and audited annually to address bias, security, and performance.
Broader implications for long-term care
While centered on veterans, the model could inform skilled nursing and long-term care beyond the VA. Industry analysts note that similar learning health approaches have cut readmissions in other settings, and the VA’s scale may accelerate adoption of data-driven quality improvement across elder care. A 2024 federal watchdog report warned that, without modernization, the VA could face a $10–15 billion shortfall in geriatric care by 2030—pressure that a learning health system aims to help offset through earlier interventions and fewer emergency visits.
The effort will be featured at the HEALeR Symposium in Indianapolis on November 6, where Regenstrief and VA teams plan to discuss applying clinical AI in learning health systems for aging populations, including links to mental health and suicide prevention goals.
For now, project leaders say the measure of success is simple: fewer avoidable hospital stays, stronger support for caregivers, and care plans that fit the realities of older veterans’ lives. Or, as Embí put it, ensuring the system “learns with and from the people it serves.”
