AdventHealth establishes an AI Advisory Board to responsibly integrate AI in healthcare. Led by Rob Purinton, the board vets AI tools based on clear principles, aligning with the health system’s mission and ensuring safety, ethics, and scalability. Physician involvement ensures practical implementation, while AI has already improved diagnostic accuracy and patient safety. AdventHealth also learns from third-party AI vendors and emphasizes effective Machine Learning Operations (MLOps) for in-house AI tools. The board seeks a balanced approach, maximizing AI’s benefits while prioritizing patient well-being.
In response to the increasing integration of artificial intelligence (AI) in healthcare, AdventHealth has taken a proactive step by creating an AI Advisory Board. This nine-state health system aims to foster in-depth discussions on the capabilities and limitations of AI. The board’s primary goal is to support clinical and IT leaders in adopting AI responsibly, ensuring that the technology is employed in a manner that prioritizes security, efficacy, ethics, and overall patient well-being.
As AI and automation continue to advance, healthcare institutions like hospitals and health systems have the opportunity to make significant strides in quality, safety, and innovation. However, the key challenge lies in effectively integrating these tools while upholding responsibility and patient safety.
Leading the AI Advisory Board is Rob Purinton, Vice President of Analytics and Performance Improvement at AdventHealth. The board comprises a diverse cross-functional team of experts, including physicians, IT specialists, data scientists, and representatives from the health system’s partner vendors such as Microsoft, Vizient, and Premier.
In a recent interview with Purinton, the reasons behind establishing the AI Advisory Board were discussed. The board aims to educate and empower clinical leaders to become advocates for the AI tools that will ultimately be implemented within AdventHealth. Given the abundance of hype, misinformation, and genuine advancements in AI, the board creates a facilitated environment where clinical leaders can navigate the complexities and support a responsible path forward for AI in healthcare.
The AI Advisory Board follows a principled approach to vetting AI tools for healthcare applications. Their evaluation process involves a funnel approach, starting with a clear problem statement and considering various AI and non-AI solutions. The selected tool must align with the health system’s mission and vision, be feasible within the existing technology framework, ensure safety, ethics, and transparency, and have the capability to scale for a large health system. Additionally, the potential workflow benefits for clinicians and the expected cost implications are taken into account, ultimately leading to higher confidence in the AI tools that pass the evaluation process.
AdventHealth has witnessed several positive impacts of AI implementation in healthcare. AI-powered tools have significantly reduced the workload for radiologists by summarizing imaging reports, enabling them to focus more on complex cases. Moreover, early detection of conditions like stroke has been made possible through AI, leading to timely interventions and improved patient outcomes. It is essential to strike a balance between AI and traditional algorithms, as both play critical roles in patient safety.
The health system has actively involved physicians in the AI vetting and implementation process. Physician leaders with extensive clinical practice experience lead and attend committees that evaluate clinical solutions, including AI tools. This involvement ensures that the decisions made align with the practical needs of healthcare providers and patients. Physician participation in vetting workgroups allows them to directly interact with vendors and data scientists, contributing to a well-informed decision-making process.
Through its journey of implementing AI tools and building in-house AI solutions, AdventHealth has continuously learned valuable lessons. When working with third-party AI vendors, they have realized the importance of having clear answers to vetting questions. As the industry moves towards standardizing AI tools, the vetting process is expected to become more streamlined for both providers and vendors. Additionally, building in-house AI tools has highlighted the significance of proper MLOps (Machine Learning Operations) management. Effective implementation involves addressing data quality, processing, and predictive accuracy concerns, and AdventHealth’s collaboration with tech vendors like Microsoft and Snowflake has contributed to continuous improvement in this area.
Overall, AdventHealth’s AI Advisory Board plays a crucial role in guiding the responsible adoption and development of AI within their healthcare system. Through thoughtful vetting processes and involving clinical leaders, AdventHealth aims to harness the potential of AI while upholding patient safety and ethical standards. As they continue to evolve and learn from their experiences, they pave the way for a more sustainable and beneficial AI-powered future in healthcare.
