GME-Naïve or “Virgin” Hospital Explorer

Purpose

This tool was created to assist Educators, Policymakers, and Workforce Planners seeking hospitals where residency expansion might be possible as a result of not having detectable prior Graduate Medical Education (GME) funding. Users will find this helpful in identifying potential “virgin” teaching hospitals, or those that never had a medical internship or residency training program.

Data & Methods

The database driving this tool was built from Fiscal Year (FY) 1996-2016 hospital cost reports from the Centers for Medicare and Medicaid Services (CMS) Healthcare Cost Report Information System.  Using these data, we identified Medicare-certified hospitals that did not have a teaching history as of FY 2016 based on a set of teaching status questions. For example, for each FY cost report, the central question asked whether the hospital was a teaching hospital (or affiliated with one) that “received payments for interns and residents.” If a hospital answered negative to this question for all years between 1996 and 2016, it was classified as a potential “virgin” hospital – a hospital with no teaching history. Teaching status was also asked for hospitals with “unapproved” teaching programs and for those that were paid on a cost basis. For those hospitals that had not yet submitted a FY 2016 cost report, we used their FY 2015 cost report as their final year. There were 6,202 unique hospitals that had submitted a cost report in FY 2015-FY2016. Out of 6,202 hospitals, 72.5% (4,495 hospitals) did not have any teaching history.

Limitations

There were limitations to our identification method: First, the teaching status questions changed between the old and the new cost report forms in early 2010s. This made it difficult for a consistent classification of teaching hospitals. Second, it was unclear whether the hospitals that did not receive any Medicare GME funding but received GME funding from other sources such as Medicaid responded to those questions. Third, we could not separately identify hospitals with a teaching program from those that had no teaching program but trained residents from teaching hospitals. Fourth, some hospitals who never claimed or were reimbursed for training residents may still be considered "non-virgin" hospitals simply by virtue of accepting residents for training (e.g. rotators). This last category would falsely be granted "virgin" status in our analysis.