Rethinking Rurality in Graduate Medical Education

Graduate medical education policy relies on rural classification to determine eligibility for rural training programs. This study aimed to assess whether hospitals classified by the Federal Office of Rural Health Policy (FORHP) as rural, but located in metropolitan counties, more closely resemble hospitals in non-metropolitan counties than other hospitals in metropolitan counties in terms of hospital and community characteristics.

This was a cross-sectional, observational study. Our study sample consisted of 4686 acute general hospitals in the United States from a 2024 publicly available hospital list. We conducted pairwise comparisons of hospital characteristics from 2024 Medicare cost report data and community characteristics from the Agency for Healthcare Research and Quality 2020 Social Determinants of Health Database and the 2024 American Medical Association Master file between FORHP-rural metropolitan hospitals and each comparison group: metropolitan non-FORHP-rural hospitals and non-metropolitan hospitals. Our main similarity analysis estimated pairwise logistic regression models predicting FORHP-rural metropolitan classification using community characteristics.

We classified 459 hospitals as FORHP-rural metropolitan hospitals. Our pairwise comparisons showed that hospital and community characteristics are similar to non-metropolitan hospital characteristics but significantly differ from metropolitan non-rural hospital characteristics. Predicted probability distributions for FORHP-rural metropolitan and non-metropolitan hospitals showed substantially greater overlap than distributions for FORHP-rural metropolitan and metropolitan non-FORHP-rural hospitals.

FORHP's tract-level rural definition identifies a distinct group of hospitals. Expanding the definition of rural could unlock federal funding for physician training for nearly 500 additional US hospitals.