Leadership of Teams Exploring AI Use in Health Care Identified as Key to Impact, With Clinicians Having Greatest Effect

Over the last decade, the use of artificial intelligence (AI) in health care has risen significantly, but AI-driven innovations in health care delivery have not fully met expectations. Research on the use of AI in health care require interdisciplinary collaboration, making team structure and leadership essential for guiding AI-driven innovation. In a new study, researchers examined how studies on AI outcomes in health care reflect team structure and leadership. They found that team leadership was significantly associated with AI impact and that teams with clinical leadership had a higher likelihood of impact.

The study was conducted by researchers at Cornell University, Bentley University, Carnegie Mellon University, the University of Maryland, New York-Presbyterian Hospital, Penn State University, Weill Cornell Medicine, Regenstrief Institute, and Indiana University Health. It is published in npj Digital Medicine.

“Health care is a complex sociotechnical system shaped by stringent regulatory constraints, inherent safety vulnerabilities, and a culture dominated by expert decision-making,” says Rema Padman, Trustees professor of management science and healthcare informatics at Carnegie Mellon’s Heinz College, who coauthored the study. “While health care delivery organizations are familiar with technological innovations, the introduction of AI differs fundamentally from previous technologies, bringing considerable challenges to health care organizations.”

These challenges stem from the multifactorial nature of AI use in health care, which requires robust clinical acceptance, technical infrastructure, and considerations for human factors. In addition, the use of AI is influenced by various organizational factors, such as technology maturity, data and AI governance policies, and leadership and interdisciplinary expertise, all of which ultimately affect its performance.

Drawing on upper echelons theory, which suggests that organizational outcomes partially reflect the values, experiences, and backgrounds of individuals in top leadership positions, researchers examined more than 100 studies conducted around the world from 2018 to 2024 on AI outcomes, most of which were randomized control trials (RCTs). Their analysis controlled for AI- and workflow-related confounders, including AI types and origin, clinical settings, and geographic region.

Leadership—that is, who led the research team and how expertise was distributed within a team—had a significant effect on AI impact, with studies led by clinicians more likely to achieve meaningful clinical impact, the researchers concluded. This finding held when looking only at RCTs and when controlling for factors such as AI type, clinical setting, AI origin, study design, team size, and authors’ organizational roles, revealing associations among type of leadership, study design, and region.
 
Specifically, when a team included a clinical champion, clinicians in that group were more likely to adopt new technologies than were members of groups without one, particularly for initial adoption, the study found. This suggests that when a team includes members with influence of the specialty, the clinicians in the team are more likely to successfully implement and use new technologies than are teams without such influential figures.

“The use of AI in health care is increasingly recognized as a complex process that extends far beyond the performance of the technology itself,” explains Yiye Zhang, associate professor of population health sciences at Cornell University, who co-led the study and received her PhD at Carnegie Mellon. “AI adoption is shaped by interdependent domains, including the nature of the condition, the technology, and its value proposition, organizational context, and broader policy environment, all of which interact dynamically over time.”

Among the study’s limitations, the authors note that their analysis was limited by publication bias and by the predominance of clinician-led studies they examined, which may have restricted underlying patterns and model results. In addition, they made assumptions about who led research teams and who contributed to the work, and they did not consider the types and characteristics of leaders in terms of job functions and leadership styles. Finally, they caution that their findings do not imply that decisions about AI use should always be made by clinicians.

“As AI transforms health care professions and job profiles through changing tasks and responsibilities in practice, organizations are starting to evolve their leadership structures,” adds Haijing Hao, associate professor of computer science at Bentley University and another co-lead, who also received her PhD at Carnegie Mellon. “Answering the question of how team leadership structure may influence outcomes of AI use can allow health care organizations to better plan for workforce development and resource planning to ensure successful use of AI in health care.”

###

Summarized from an article in npj Digital Medicine, The Impact of Leadership on AI Deployment Study Outcomes in Healthcare: An Integrative Analysis, by Li, Q (University of Maryland), Li, P (Cornell University, New York-Presbyterian Hospital), Hao, H (Bentley University), Padman, R (Carnegie Mellon University), Gao, H (Penn State University), Gossey, JT (Weill Cornell Medicine), Bian, J (Regenstrief Institute, Indiana University Health), and Zhang, Y (Cornell University, New York-Presbyterian Hospital, Weill Cornell Medicine). Copyright 2025 The Authors. All rights reserved.

About Heinz College of Information Systems and Public Policy
The Heinz College of Information Systems and Public Policy is home to two internationally recognized schools: the School of Information Systems and Management and the School of Public Policy and Management. Heinz College leads at the intersection of people, policy, and technology, with expertise in analytics, artificial intelligence, arts & entertainment, cybersecurity, health care, and public policy. The college offers top-ranked undergraduate, graduate, and executive education certificates in these areas. Our programs are ranked #1 in Information Systems, #1 in Information and Technology Management, #8 in Public Policy Analysis, and #1 in Cybersecurity by U.S. News & World Report. For more information, visit www.heinz.cmu.edu.