UNC Health’s CIO, Brent Lamm, is at the forefront of AI advancements in healthcare. Working with Epic and Microsoft, Lamm is involved in a pilot project to enhance the provider experience using AI capabilities. They aim to reduce clinician burnout and improve patient-provider relationships. Lamm emphasizes the need for careful implementation, cautioning providers about the limitations of AI technology. He also highlights the collaboration with Microsoft to leverage generative AI for various use cases, such as assisting care teams in accessing training resources efficiently.
UNC Health’s Chief Information Officer (CIO), Brent Lamm, is at the forefront of the healthcare industry’s artificial intelligence (AI) revolution. With his extensive knowledge and experience, Lamm is eager to share insights and advice with his peers.
Based in Morrisville, North Carolina, Lamm holds a key position in the ongoing AI advancements in healthcare. He is involved in a pilot project led by Epic, a leading electronic health records company, in collaboration with Stanford, the University of Wisconsin, and UC San Diego. The project aims to test Epic’s large language model AI capabilities to enhance the provider experience.
Lamm is also engaged in exciting AI work with Microsoft, specifically focusing on generative AI. This technology, similar to the one behind applications like ChatGPT, is being applied to various use cases to support clinicians and other team members.
In a recent interview with Healthcare IT News, Lamm delved into these projects and shared his perspectives on the growing role of AI in healthcare.
Q: Can you tell us about the goals of the pilot project you’re working on with Epic and other institutions?
A: Our primary objectives are to combat clinician burnout and enhance the patient-provider relationship through the utilization of AI technology for drafting responses to patient communications.
While the past decade has witnessed significant benefits from digital solutions enabling easier communication between patients and their care teams, it has also placed an increased workload on clinicians. This has led to challenges in maintaining the personal well-being of care team members, which were further exacerbated by the pandemic and increased patient reliance on digital communication.
We believe that the generative AI capabilities being developed by Epic can play a crucial role in addressing these challenges by reducing the cognitive burden associated with drafting responses. This technology functions similarly to how spreadsheet software has relieved business users of basic calculations, enabling them to focus on core problems or opportunities.
By offloading the task of crafting draft communications, clinicians can better concentrate on the needs of their patients. This more efficient process should lead to an improved patient experience and a stronger relationship with the care team.
As early adopters of generative AI within our Epic EHR system, we aim to expedite this transformative process while earning the trust of our care teams by actively participating in the development of an effective end product.
We hope that this initial pilot project will be the first of many Epic AI initiatives supported by our team. We are enthusiastic about the potential benefits and efficiency that AI can offer our care team members and administrative Epic users.
Q: Could you elaborate on how you and your team will use the large language model AI to enhance the EHR experience for providers? How will you help providers understand AI technology?
A: Initially, we will introduce this new capability to approximately ten physicians in a controlled environment to ensure safe testing of the functionality. Our team comprises physicians from various specialties, including primary care, who will evaluate the usability and accuracy of the communications drafted by the generative AI.
One important metric we will assess is the effectiveness of the technology in reducing the time required for providers to finalize their communications. The amount of editing needed will be a crucial measure of success. Ultimately, we aim to assist Epic in refining the technology to a point where minimal edits are necessary.
Another significant aspect of our work involves evaluating provider satisfaction with the tone and style of the draft messages. The generated content must align with their expectations and preferred communication methods with patients. This aspect holds significant promise from a technological and informatics standpoint.
Furthermore, our team is excited about evaluating and contributing to the development of generative AI in specialty-specific contexts. We want to ensure that this technology is accessible and effective for all of our clinicians.
Q: Considering the concerns surrounding AI, what cautions will you provide to your providers regarding the large language model of AI?
A: We are pleased to see that Epic is taking a thoughtful and cautious approach to implementing this technology. One of our key responsibilities is to adequately train and educate the providers participating in this program.
The primary caution is to ensure a thorough review of the drafted message before sending it. Generative AI is currently far from perfect and can produce inaccurate or biased text, sometimes referred to as “hallucination.” While this issue will improve over time as the AI learns, users must carefully validate automatically generated drafts. Everyone needs to understand that this technology is merely a tool to aid in drafting the initial message, with providers retaining full control over the final response.
In parallel with this effort, we have launched a formal program involving leaders from our organization to develop and implement UNC Health’s Responsible AI Framework. This framework, led by a committee of multidisciplinary leaders and clinicians, will evaluate both vendor-developed and in-house AI solutions, considering factors such as fairness, transparency, accountability, and trustworthiness.
Looking ahead, as AI technology continues to advance, it is crucial to maintain the perspective of AI as “augmented intelligence” and uphold it as a core principle. While machine learning, large language models, and other forms of AI have the potential to revolutionize healthcare positively, they should be viewed as tools to support our most valuable asset: our people.
Q: You’re also collaborating with Microsoft on an AI project. Can you provide more details about this initiative and its objectives?
A: Yes, we are thrilled to be closely collaborating with Microsoft to leverage generative AI for a wider range of use cases to support our clinicians and other team members. Although the project is still in its early stages, we have already developed a fully functional prototype of an internal general-purpose chatbot, similar to the consumer-facing application ChatGPT.
We believe that a local, enterprise solution like this has the potential to unlock the productivity and efficiency gains that generative AI chatbots offer while ensuring safety, security, and privacy for our users.
One of our initial focus areas for leveraging this capability is to assist our care teams in accessing training and education resources more quickly and easily. Given the ever-evolving protocols and workflows, frontline care team members frequently refer to these materials. We believe that chatbot-style search functionality can significantly improve efficiency, enabling our users to find the specific information they need swiftly and subsequently shift their focus back to patient care.
Q: What advice would you give to your CIO peers at other healthcare provider organizations regarding AI implementation today?
A: First and foremost, we must keep the patient and our care team members at the core of our AI initiatives. This emerging technology has immense potential to improve health outcomes, reduce clinician burnout, and positively impact healthcare costs.
However, we must never lose sight of our mission to enhance the health and well-being of the patients and communities we serve.
While it may sound paradoxical, I strongly encourage industry leaders to aggressively pursue the use of AI, but carefully and thoughtfully. It is also crucial to find trusted partners who share our values concerning patient-centered care and data privacy protection.
Given the nature of this technology and the substantial resources required for its implementation, most organizations will need to align with strategic partners. Ensuring alignment of values is equally, if not more, important than aligning technology and architecture.
I eagerly await witnessing how our peer colleagues and organizations embrace and drive improved outcomes with AI. When approached correctly, the possibilities are truly astounding.