Leveraging AI-enabled learning health systems to advance value-based health care: A conceptual framework

Despite rapid innovation, health care systems face a persistent 17-year gap between evidence discovery and implementation, undermining efforts to deliver value-based care. Bridging this "know-do gap" is essential to improving outcomes and reducing waste. Existing Learning Health System (LHS) frameworks often lack mechanisms to institutionalize learning at speed and scale.

We propose an AI-enabled LHS framework that leverages artificial intelligence (AI) to connect micro-level clinical learning with macro-level organizational decision-making. Grounded in organizational learning theory, our model illustrates how AI accelerates knowledge capture, conversion, and institutionalization via continuous, bidirectional feedback loops. AI enables real-time learning cycles, linking patient-provider data ("micro") to system-wide insights and policy adjustments ("macro"), and back to point-of-care decision support.

Our framework advances the LHS paradigm by adding speed, scale, and micro↔macro integration. Unlike earlier models, it centers AI not as an adjunct but as a foundational learning engine. Case examples from UCHealth and Mass General Brigham show how AI can drive real-time operational learning and institutional memory through structured governance and data infrastructure.

To implement an AI-LHS, organizations should (1) assess readiness and align on value-based goals; (2) invest in data infrastructure and interoperability; (3) cultivate a learning culture by engaging clinicians and staff; (4) embed AI into continuous improvement cycles with interdisciplinary governance; (5) adopt a sociotechnical approach integrating people, processes, and technology; and (6) ensure safeguards for equity, privacy, and security. These steps allow systems to reduce lag between insight and impact, accelerating value-based care transformation.