Importance: Graduate medical education (GME) funding consists of more than $10 billion annual subsidies awarded to academic hospitals to offset the cost of resident training. Critics have questioned the utility of these subsidies and accountability of recipient hospitals.
Objective: To determine the association of GME funding with hospital performance by examining 3 domains of hospital operations: financial standing, clinical outcomes, and resident academic performance.
Design, setting, and participants: This study is an economic evaluation of all academic centers that received GME funding in 2017. GME funding data were acquired from the Hospital Compare Database. Statistical analysis was performed from May 2016 to April 2020.
Exposures: GME funding.
Main outcomes and measures: This study assessed the association between GME funding and each aspect of hospital operations. Publicly available hospital financial data were used to calculate a financial performance score from 0 to 100 for each hospital. Clinical outcomes were defined as 30-day mortality, readmission, and complication rates for a set of predefined conditions. Resident academic performance was determined by Board Certification Examination (BCE) pass rates at 0, 2, and 5 years after GME funding was awarded. Confounder-adjusted linear regression models were used to test association between GME funding data and a hospital's financial standing, clinical outcomes, and resident academic performance.
Results: The sample consisted of 1298 GME-funded hospitals, with a median (IQR) of 265 (168-415) beds and 32 (10-101) residents per training site. GME funding was negatively correlated with hospitals' financial scores (β = -7.9; 95% CI, -10.9 to -4.8, P = .001). Each additional $1 million in GME funding was associated with lower 30-day mortality from myocardial infarction (-2.34%; 95% CI, -3.59% to -1.08%, P < .001), heart failure (-2.59%; 95% CI, -3.93% to -1.24%, P < .001), pneumonia (-2.20%; 95% CI, -3.99% to -0.40%, P = .02), chronic obstructive pulmonary disease ( -1.20%; 95% CI, -2.35% to -0.05%, P = .04), and stroke (-3.40%; 95% CI, -5.46% to -1.33%, P = .001). There was no association between GME funding and readmission rates. There was an association between higher GME funding and higher internal medicine BCE pass rates (0.066% [95% CI, 0.033% to 0.099%] per $1 million in GME funding; P < .001).
Conclusions and relevance: This study found a negative linear correlation between GME funding and patient mortality and a positive correlation between GME funding and resident BCE pass rates in adjusted regression models. The findings also suggest that hospitals that receive more GME funding are not more financially stable.