AI is solving the wrong physician shortage problem, doctor says
NEW YORK, UNITED STATES — A practicing neurocritical care physician and neurosurgery professor argues the United States is misdiagnosing its physician shortage — and that AI tools targeting productivity miss the underlying problem entirely, Medcity News reports.
Physician shortage misdiagnosed as supply crisis
Dr. Marc Ayoub, founder of Saile and assistant professor of neurosurgery at the Donald & Barbara Zucker School of Medicine, says the framing of a physician shortage obscures the real issue.
“We are not running out of doctors. We are running a system that is remarkably effective at preventing qualified physicians from doing the work they trained for,” Ayoub said.
McKinsey’s U.S. Physician Survey found approximately 35% of physicians report a likelihood of leaving their current roles within five years.
Of those, roughly 60% expect to exit clinical practice entirely — not retire, but leave medicine while still in their prime working years.
Forty to 50% of physicians are already picking up additional work outside their primary roles, according to Forbes reporting — suggesting available capacity exists but is not being deployed inside the system.
Fixing workflow beats recruiting new doctors
Ayoub’s argument points directly at operational friction as the cause — not a shortage of trained physicians.
“This isn’t a supply problem. It’s a distribution problem. The physicians exist. The system just isn’t getting them where they need to be,” he said.
A Deloitte survey found 64% of physicians identified provider credentialing as a major workflow improvement opportunity, alongside prior authorizations and pharmacist communications.
McKinsey data adds another layer. Fifty-one percent of physicians identify schedule flexibility as a key retention factor, yet only 59% of workplaces actively pursue such enhancements.
“Getting the diagnosis right is the first step toward the right treatment. We do not need more doctors as urgently as we need the ones we already have to be able to work,” Ayoub said.
Health systems investing in AI tools that reduce documentation burden are addressing one piece of the problem. But prior authorization processing, credentialing workflows, and scheduling infrastructure are not tasks AI productivity tools reach.
The healthcare outsourcing sector — a multibillion-dollar industry covering prior authorization support, revenue cycle management, medical coding, and clinical documentation — handles exactly the administrative load Ayoub identifies as the system’s central bottleneck.
For health system executives confronting a 35% physician departure risk, the intervention that keeps those physicians in practice may be operational, not technological. The answer is in the bottleneck, not in the pipeline.
