STAT+: NEJM AI’s editor-in-chief on FDA approval vs. clinical effectiveness
In this edition of AI Prognosis: A chat with the editor of New England Journal of Medicine's new journal focused on AI.
You’re reading the web edition of STAT’s AI Prognosis newsletter, our subscriber-exclusive guide to artificial intelligence in health care and medicine. Sign up to get it delivered in your inbox every Wednesday.
Isaac “Zak” Kohane is the chair of biomedical informatics at Harvard Medical School. He’s uniquely situated for the explosion of AI in health care, having a Ph.D. in computer science as well as an M.D. He’s also been a proponent of using clinical trials to test the effectiveness of AI in medicine.
Last fall, he proposed in NEJM something I thought was interesting: Testing ChatGPT’s advice not against doctor’s advice, but against the real-world alternative to asking Dr. Google or Dr. ChatGPT: Not going to the doctor at all. However, that comes with trade-offs, as we saw a couple weeks ago — not including the best possible control group means we miss out on learning where AI’s advantages are compared to other methods.
