In a 2023 study published in the Annals of Emergency Medicine, European researchers fed the AI system ChatGPT information on 30 ER patients. Details included physician notes on the patients’ symptoms, physical exams, and lab results. ChatGPT made the correct diagnosis in 97% of patients compared to 87% for human doctors.
AI 1, Physicians 0
JAMA Cardiology reported in 2021 that an AI trained on nearly a million ECGs performed comparably to or exceeded cardiologist clinical diagnoses and the MUSE (GE Healthcare) system›s automated ECG analysis for most diagnostic classes.
AI 2, Physicians 0
Google’s medically focused AI model (Med-PaLM2) scored 85%+ when answering US Medical Licensing Examination–style questions. That›s an «expert» physician level and far beyond the accuracy threshold needed to pass the actual exam.
AI 3, Physicians 0
A new AI tool that uses an online finger-tapping test outperformed primary care physicians when assessing the severity of Parkinson’s disease.
AI 4, Physicians 0
JAMA Ophthalmology reported in 2024 that a chatbot outperformed glaucoma specialists and matched retina specialists in diagnostic and treatment accuracy.
AI 5, Physicians 0
Should we stop? Because we could go on. In the last few years, these AI vs Physician studies have proliferated, and guess who’s winning?
65% of Doctors are Concerned
Now, the standard answer with anything AI-and-Medicine goes something like this: AI is coming, and it will be a transformative tool for physicians and improve patient care.
But the underlying unanswered question is:
The Medscape 2023 Physician and AI Report surveyed 1043 US physicians about their views on AI. In total, 65% are concerned about AI making diagnosis and treatment decisions, but 56% are enthusiastic about having it as an adjunct.
Cardiologists, anesthesiologists, and radiologists are most enthusiastic about AI, whereas family physicians and pediatricians are the least enthusiastic.
To get a more personal view of how physicians and other healthcare professionals are feeling about this transformative tech, I spoke with a variety of practicing doctors, a psychotherapist, and a third-year Harvard Medical School student.
‘Abysmally Poor Understanding’
Alfredo A. Sadun, MD, PhD, has been a neuro-ophthalmologist for nearly 50 years. A graduate of MIT and vice-chair of ophthalmology at UCLA, he’s long been fascinated by AI’s march into medicine. He’s watched it accomplish things that no ophthalmologist can do, such as identify gender, age, and risk for heart attack and stroke from retinal scans. But he doesn›t see the same level of interest and comprehension among the medical community.
“There’s still an abysmally poor understanding of AI among physicians in general,” he said. “It’s striking because these are intelligent, well-educated people. But we tend to draw conclusions based on what we’re familiar with, and most doctors’ experience with computers involves EHRs [electronic health records] and administrative garbage. It’s the reason they’re burning out.”
Easing the Burden
Anthony Philippakis, MD, PhD, left his cardiology practice in 2015 to become the chief data officer at the Broad Institute of MIT and Harvard. While there, he helped develop an AI-based method for identifying patients at risk for atrial fibrillation. Now, he’s a general partner at Google Ventures with the goal of bridging the gap between data sciences and medicine. His perspective on AI is unique, given that he’s seen the issue from both sides.