How UNC Health, Epic, and Microsoft Leverage Generative AI For Healthcare

TL;DR Breakdown

• UNC Health’s CIO is leading the way in using generative AI from Epic and Microsoft to reduce clinician burnout and enhance patient-provider relationships. 
• The pilot project aims to automate drafting patient communications, allowing clinicians to focus more on patient needs and improving overall experiences. 
• UNC Health emphasizes caution and proper training when using AI, maintaining a responsible approach, and ensuring patient-centered care. 

Brent Lamm, the CIO at UNC Health, is at the forefront of the healthcare artificial intelligence (AI) revolution. Working closely with Epic and Microsoft, Lamm and his team pilot projects that leverage AI capabilities to enhance the provider experience and transform healthcare delivery. In an interview with Healthcare IT News, Lamm shares insights into the goals, implementation, and cautious approach to AI adoption, emphasizing the importance of patient-centered care and collaboration with strategic partners.

The pilot project’s goal involving Epic, Stanford, the University of Wisconsin, and UC San Diego is to utilize Epic’s large language model AI capabilities to reduce clinician burnout and enhance the patient-provider relationship. With the increased adoption of digital communication tools, clinicians face a higher workload, negatively impacting their well-being. The project aims to alleviate this burden by developing generative AI capabilities that assist clinicians in drafting responses to patient communications. Automating this process allows clinicians to focus more on patients’ needs, leading to improved experiences and stronger relationships.

Implementing LLM AI to improve EHR experience for providers

UNC Health’s initial work with LLM AI involves exposing approximately ten physicians to the new functionality in a controlled environment. These physicians, from various specialties, including primary care, will evaluate the usability and accuracy of AI-generated communications. Key metrics will include the time reduction in finalizing communications and the required editing level. The team also aims to ensure the generated content aligns with providers’ communication preferences and expectations. Furthermore, UNC Health intends to contribute to the development of generative AI, considering specialty-specific differences and nuances to make the technology effective for all clinicians.

Lamm emphasizes the need for caution and proper training when utilizing LLM AI. While the technology is imperfect and may generate inaccurate or biased content, providers should carefully review and validate drafted messages before sending them. UNC Health has launched a formal program, the Responsible AI Framework, to evaluate AI solutions based on fairness, transparency, accountability, and trustworthiness considerations. Lamm believes that maintaining the perspective of AI as “augmented intelligence” is crucial, using AI as a tool to support healthcare professionals rather than replacing them.

Microsoft partnership and the future of AI in healthcare

UNC Health collaborates with Microsoft to leverage generative AI in various use cases. They have developed a fully functional prototype of an internal chatbot, similar to ChatGPT, to assist clinicians and team members. The initial focus is on helping care teams access training and education resources more efficiently. Using chatbot-style search, UNC Health aims to provide quick and accurate access to specific information, enabling clinicians to focus more on patient care. The project aligns with the organization’s commitment to maintaining its users’ safety, security, and privacy.

Advice to CIO Peers and pursuing AI with care and thoughtfulness

Lamm advises CIO peers in healthcare organizations to pursue AI adoption while remaining cautious and thoughtful aggressively. He emphasizes the importance of keeping patients and care team members at the center of AI advancements and prioritizing patient-centered care and data privacy. Strategic partnerships with shared values become crucial in implementing AI successfully. Lamm looks forward to seeing how colleagues and organizations embrace AI to drive improved outcomes, highlighting AI’s tremendous possibilities when deployed responsibly and effectively.

Brent Lamm’s work at UNC Health exemplifies the transformative potential of AI in healthcare. Through collaborations with Epic and Microsoft, UNC Health is leading initiatives to reduce clinician burnout, enhance patient-provider relationships, and unlock productivity gains using generative AI capabilities. With a responsible and cautious approach, Lamm emphasizes the importance of maintaining patient-centered care and aligning with partners who share the organization’s values. By embracing AI with care and thoughtfulness, healthcare organizations can revolutionize healthcare delivery and improve health outcomes for patients and communities.