The Veterans Affairs Department launches a million-dollar AI challenge to devise data management and clinical practice enhancements for reducing clinician burnout, inviting innovators to collaborate in a tech sprint for improved healthcare delivery to veterans. Applications for the ambitious project close in January 2024, with substantial monetary rewards for the top solutions.
The Department of Veterans Affairs (VA) has issued a call to action for healthcare innovators to design AI tools to improve the handling of healthcare data and clinical operations.
Aiming to alleviate the strain on healthcare professionals, the VA has launched a 120-day initiative, known as a Tech Sprint, focused on creating seamless dictation for clinical consultations and a more efficient processing system for the Community Care program documentation. This new system must be adept at generating electronic health record-friendly data.
The Significance
The VA’s Discovery, Education, and Affiliate Networks office, in collaboration with the National Artificial Intelligence Institute, SimLEARN, and the Office of Primary Care, announced this AI challenge. This announcement aligns with the recent AI Executive Order by the Biden Administration.
The challenge, titled “AI Tech Sprint for Documenting VA Clinical Encounters and Integrating Community Care Data,” invites academic entities, industry partners, and non-profits to join forces with VA professionals and contributors to develop solutions to mitigate clinician burnout.
The push to reduce burnout is urgent for the VA, as it operates the nation’s most extensive healthcare network, providing education to a significant number of the country’s nurses and doctors.
In the wake of recruiting at unprecedented levels, the VA underscores the importance of reliable AI technology to meet the growing demand for veteran care and services. The VA emphasizes the requirement for precise and accountable documentation of healthcare interactions, which should be compatible with the VA’s health information systems. This technology must uphold the highest security standards and produce summaries for patients, while aligning with clinical documentation norms.
Another aspect of this initiative is geared towards enhancing documentation for Community Care, which is healthcare offered to veterans by external providers, under VA sponsorship. The sought-after AI system should be adept at assimilating varied Community Care documentation, making important care episodes easily identifiable and retrievable.
The designed system should fit into the VA’s CDW Delta Lake and assimilate various data sources, employing sophisticated entity recognition and summarization of medical texts, in line with VA’s technical and monitoring standards.
The system must also be able to accurately source and process structured data into the VA’s electronic health records and Summit Data Platform Health Information Exchange, including CPT, SNOMED CT, and LOINC codes.
The submission window closes on January 5, 2024, with the sprint commencing on January 26. Prizes of $300,000 for first place, $150,000 for second, and $50,000 for third are on offer for both challenges.
Broader Context
Previously, the VA has funded AI initiatives to prevent veteran suicides, with notable projects like ReflexAI, the Battle Buddy app, and Stop Soldier Suicide’s Black Box Project making strides in crisis response and risk identification.
The latest tech sprint is part of an overarching directive from President Biden on Safe, Secure, and Trustworthy AI, addressing the potential impacts of AI on the labor market and advocating for responsible AI innovation.
The VA undersecretary for health, Dr. Shereef Elnahal, has stated on record, “Implementing AI solutions can free up our clinicians from administrative tasks, allowing them to focus on their passion for veteran care. This advancement is key to both reducing clinician burnout and elevating the level of healthcare for veterans.”
