In a recent episode of TEDMED Conversations, Professor Anupam Jena of Harvard Medical School's Department of Health Care Policy discussed the emerging role of Artificial Intelligence(AI) in medicine with host Kelly Thomas, PhD. Drawing on his expertise as an economist, physician, and research professor at HCP, Jena highlighted two transformative ways AI is revolutionizing medicine: enhanced decision-making and the development of new medical insights.
AI acts as a vigilant and informed assistant, empowering medical professionals to make well-rounded decisions, making it an invaluable resource in emergency situations where practitioners may be juggling factors which distract from ordinary information accessibility. By offering real-time analysis of diagnostic tests, such as EKGs, in emergency situations, AI can quickly identify missed diagnoses or hidden patterns. Additionally, AI can tirelessly scan electronic health records, alerting physicians to emerging concerns. This background monitoring results in improved care delivery by providing a comprehensive view of patient data, guiding care plans, and optimizing treatment strategies.
Jena expressed interest in AI's ability to analyze vast amounts of medical data enabling understanding in ways never before possible. While traditional diagnosis relies on recognition of established patterns, AI-powered pattern recognition can identify subtle patterns not yet widely recognized by humans, leading to discovery of patterns previously overlooked or unseen. AI can detect diseases in their earliest stages and identify previously unknown indicators of disease. Such advanced pattern recognition is a powerful new tool for medical discovery which will result in better patient outcomes.
When asked about the role of AI in natural experiments, Jena expanded on how natural experiments are effective in leveraging real-world data analysis of a higher volume of patient cases, in developing scenarios, free from the constraints of randomized controlled trials (RCTs).
While RCTs are the gold standard for establishing cause and effect between drugs and outcomes, they can be expensive and time-consuming. And while real-world data provides insights into how medications work in everyday practice, simply analyzing data on who takes a drug, and the outcome, can be misleading due to various influencing factors.
However, natural experiments are free from the limitations of traditional trials and offer in the moment opportunities for discovery, utilizing the vast potential of real-world data allowing for analysis of a higher volume of patient cases to improve medical knowledge and practice.
By utilizing real-world situations where different patients receive unique treatments due to factors outside of the study design, such as hospital drug shortage, or physician prescription preference, natural experiments allow researchers to observe the effects of treatments without the practical limitations of RCTs. Analysis of natural experiment outcomes can provide valuable insights into "what works" and "what doesn't work" in medicine, leading to better treatment strategies and improved patient care.
Jena’s research involves several areas of health economics and policy including the use of natural experiments in health care, the economics of physician behavior and the physician workforce, medical malpractice, the economics of health care productivity, and the economics of medical innovation. He is the host of the Freakonomics, MD podcast, which explores the “hidden side of health care.”
Want to learn more about Dr. Jena's work?Check out his recent book, Random Acts of Medicine, co-authored with Dr. Christopher Worsham.