Open Health Data Hub



All posts
|3 min read

Foreign Medical Schools Now Train a Surprising Share of U.S. Specialists

international medical graduatesforeign medical schoolphysician workforceMedicare clinician trainingIMG doctors

Graduates of non-top-50 medical schools now outnumber their top-50 counterparts in every major U.S. specialty examined. That's not a rounding error or a quirk of one underserved field. It holds across Family Practice, Psychiatry, Internal Medicine, and Nephrology.

The assumption that America's physician workforce flows primarily from a recognizable tier of large, well-resourced domestic schools is simply wrong.

The Majority Rule Holds Across Every Specialty Examined

SpecialtyTotal CliniciansNon-Top-50 Share
Family Practice56,67755.8%
Psychiatry13,41654.5%
Internal Medicine47,05553.5%
Nephrology3,94851.3%

The range is narrow: 51.3% to 55.8%. That consistency across four structurally different specialties, from a high-volume primary care field to a relatively small subspecialty, makes the pattern harder to dismiss as coincidental. This isn't a Family Practice story or a primary care story. It's a workforce story.

Family Practice sits at the top of the range, with 55.8% of its 56,677 clinicians trained outside the top 50 U.S. institutions by enrollment. Top-50 graduates represent only 44.2% of the specialty, the lowest share among the four. For a field already under pressure from demand outpacing supply, the practical implication is direct: the majority of the clinicians filling that gap did not come from the schools most people picture when they think about physician training.

Even Subspecialties Depend on This Pipeline

Nephrology is the most competitive case for the traditional assumption. It has the highest top-50 graduate share of the four specialties, at 48.7%. Yet even there, non-top-50 graduates constitute the majority at 51.3% of 3,948 clinicians.

Nephrology is a demanding subspecialty requiring additional fellowship training after internal medicine residency. The fact that graduates of smaller or foreign institutions make up the majority even in this field undercuts the idea that elite domestic training is a prerequisite for complex specialty practice. Patients receiving dialysis or transplant care are, more often than not, being treated by a physician who trained outside the top 50.

Internal Medicine reinforces the point at scale. With 47,055 total clinicians, it's the largest specialty in this analysis, and 53.5% of those clinicians attended non-top-50 schools. Internal medicine serves as the backbone of hospital medicine and feeds into nearly every subspecialty pipeline. A majority-non-top-50 workforce at that level shapes the composition of the entire specialist tier above it.

What "Non-Top-50" Actually Means for the Workforce

The category "non-top-50 by enrollment" includes both smaller domestic medical schools and international medical schools. The Clinician Directory data doesn't separate those two groups within this analysis, which matters. A graduate of a mid-sized state medical school and a graduate of a Caribbean or South Asian medical school both fall into the same bucket here.

That distinction has real consequences for how workforce planners think about pipeline resilience. If a substantial portion of the non-top-50 majority trained abroad, then U.S. specialty care is more dependent on international medical graduates than most domestic training capacity discussions acknowledge. If the majority trained at smaller domestic schools, the story is about geographic and institutional diversity within the U.S. system. The aggregate numbers can't resolve that question, but they do establish that the top-50 pipeline alone cannot account for the majority of practicing clinicians in any of these four fields.

Across all four specialties, more than half of active clinicians came from outside the institutions most commonly associated with physician production. Workforce models built around expanding capacity at top-tier schools are, by definition, addressing a minority of the current supply chain.

Explore the data yourself

Run your own queries against 240M+ rows of federal health data using natural language — powered by AI.

Start analyzing