The 2014 GME Residency Match Results: Is There Really A “GME Squeeze”?

Source: Health Affairs Blog

By Edward Salsberg

Each spring thousands of seniors at medical and osteopathic schools and other physicians apply for positions in graduate medical education (GME) training programs; simultaneously, thousands of training programs rank their preferred candidates.  Based on the preferences of the medical student/physician applicants and the training programs, the two are matched by a sophisticated computer program.  Since GME is a prerequisite to becoming licensed and practicing medicine in the US, this is a critical juncture in the education – training pipeline and provides a spotlight on the future physician workforce.

There are two matching systems: one administered by the National Residency Match Program (NRMP) for allopathic training positions, accredited by the Accreditation Council for Graduate Medical Education (ACGME), that matches medical doctors (MDs), doctors of osteopathic medicine (DOs) and graduates of schools outside of the US, known as international medical school graduates (IMGs); and one for GME programs accredited by the American Osteopathic Association (AOA) that is limited to DOs.  The following are among the highlights of the results of this year’s matches.

First year positions (PGY 1 positions) for entrants into GME reached an all-time high and the number continues to grow. This year, a record 26,678 first year positions were offered by the NRMP and an additional 2,988 first year positions were offered in the AOA sponsored match, for a total of 29,666 positions offered in 2014. (See Note 1) This represents an overall increase of 2.2 percent from 2013. (See Note 2)  However, some of the NRMP increase may reflect the “all in” policy instituted by the NRMP effective in 2013. (See Note 3)

Entry level GME positions far outnumber the number of US medical and osteopathic graduates seeking a residency position.  Despite a lot of rhetoric and fear that new US graduates are facing a lack of training slots, overall, there were about 22,300 US MD and DO seniors competing for the 29,666 first year positions.

The number of graduates of foreign medical schools continues to be significant. More than 6,350 IMGs were matched into ACGME programs. (IMGs are not eligible for AOA accredited training programs.)  Between 2013 and 2014, there was a very small increase in the number of IMGs matched through the NRMP.  However, the NRMP “all in” policy had the greatest impact on IMGs participating in the match, making comparisons prior to 2013 match data inappropriate, and some of the 2014 increase may reflect the residual impact of that change.  Of the IMGs matched through the NRMP in 2014, 2,722 (43 percent) were US citizens who had graduated from foreign medical schools.

Some MD seniors were not matched in the main match.  For MDs, 975 of 17,374 active senior participants were not matched in the main match.  This was down 11 percent from 1,097 in 2013.  There is a supplemental match process, the Supplemental Offer and Acceptance Program (SOAP), which allows unmatched applicants to try to obtain an unfilled residency position during the few days following the main match (previously known as the “scramble”). In 2014, there were 991 entry level positions unmatched in the NRMP main match. While this is very close to the 975 senior MDs who were not matched, several thousand unmatched DOs, IMGs and other US MDs (non-seniors) were also competing for those 991 unfilled positions. While 1,097 senior MDs were unmatched after the main match in 2013, this was reduced to 526 US MD seniors (3 percent of all seniors) after the supplemental match.  (Those that remained unmatched were eligible to apply again in 2014.)

Some DO graduates did not get matched in the main matches. The AOA sponsored match occurs prior to the NRMP match. As a result, the figure for osteopathic seniors who do not match is less clear, as they can apply to ACGME programs through the NRMP. A total of 611 DOs that participated in the NRMP were unmatched. They too could compete for the 991 positions in the NRMP supplemental match.

There appears to be a slight increase in the numbers of physicians going into primary care.  One bright spot is a steady increase in both NRMP and AOA matches of physicians going into family medicine; the combined increase was 3.2 percent between 2013 and 2014 and it was driven by an increase in matches for US MD and DO seniors.  For both match programs this seems to be a continuation of a steady increase over the past 5 years.  This is encouraging because about 95 percent of residents entering family medicine training actually end up practicing primary care.  Among all US MD seniors in the match, 8.5 percent were matched into family medicine; representing a slight increase over prior years.  However, US MDs only comprised 45 percent of the NRMP matches in family medicine in 2014.  There were also increases in NRMP and AOA matches in internal medicine and pediatrics, both of which lead to primary care; but this does not provide a clear picture of the future primary care workforce, as many or even most of these physicians will go on to sub-specialize or become hospitalists. Data from a point later in the education-training pipeline is needed to assess if the percent of internists and pediatricians going into primary care is changing.


The “GME squeeze.”  A number of groups have suggested that the nation needs to expand GME funding in order to increase GME slots to assure that all medical and osteopathic school graduates are able to complete the training required to become a licensed physician. The argument is that the number of medical and osteopathic graduates is growing faster than GME slots and that eventually there will be more graduates than entry level residency positions.  Based on the match data, it is clear that there are currently far more GME entry positions than graduates.  Analysis shows that using the recent growth rates of graduates and GME slots, there will continue to be several thousand more GME positions than graduates at least for the next decade.  Thus, while finding a GME position may be getting more competitive, in no way is it a tight squeeze.

Nevertheless, we should be concerned with the fact that several hundred MD and DO graduates were not able to find a GME position. Adding GME positions will not assure positions for all US seniors.  The issue needs to be examined closely to identify the root causes of the problem.  Issues to be explored include the following:

  1. Does this reflect poor medical student decision making around specialty choice? For example,according to the NRMP, in 2013, 724 US seniors applied only for positions in orthopedic surgery; however, there were only 693 orthopedic positions available altogether and a few went to applicants other than US seniors.  Would better student counseling help?
  2. Do some GME programs prefer foreign medical school graduates? If so, why?  Given the rigorous accreditation process for US medical education, it is appropriate to ask why so many graduates of foreign medical schools were accepted when US MD and DO grads were unmatched. Is this something the ACGME should explore?
  3. Are some US graduating students inadequately prepared for GME? Do medical and osteopathic schools need to review their curriculum?
  4. Would a modification in the current NRMP match process, such as allowing a two-stage process with the most competitive specialties going first, better serve programs and seniors?

Foreign medical school graduates.  The number of IMGs matched through the NRMP increased slightly from 6,247 in 2013 to 6,355 in 2014, an increase of 1.7 percent. Some of this increase could reflect additional programs entering the NRMP.  Regardless, 6,335 IMGs is a significant number representing 24.7 percent of all NRMP matches and 22.9 percent of all NRMP and AOA year matches combined. This is well above the percent of foreign-educated individuals in most other health professions, like nursing and pharmacy, which are generally in the 5 to 10 percent range.  The primary beneficiary of increasing GME slots would be IMGs, who are facing the greatest pressure from the increasing competitiveness for GME positions.  Increasing the number of IMGs would be contrary to theWorld Health Organization’s Code of Practice on the International Recruitment of Health Personnel, which was signed by the US and 182 other countries. The Code called for countries to do a better job of meeting their own needs rather than relying on foreign educated health professionals.

Specialty choice.  As noted above, while the increase in the numbers entering family medicine in both ACGME and AOA accredited programs is encouraging, the match data on entrants into GME in general do not provide a clear picture of the future primary care physician workforce.  Since this is a critical issue for the health care system, more needs to be done to develop and monitor data later in the training process and post training to assess the trends in primary care.  One particular challenge is getting accurate and timely data on the numbers of internists and pediatricians becoming hospitalists.  Major gaps currently exist in this area.

Note 1. Despite the NRMP “all in” policy (described below) a few programs may still match outside of the formal match system, which would lead to an under count of total new entrants. However, there are some programs that are dually certified by the ACGME and the AOA. This could lead to a slight over count of positions. Both of these phenomena seem to be relatively limited in 2014.

Note 2. This is consistent with earlier findings.  Jolly and coauthors found 0.9 percent annual growth of ACGME positions; while an earlier study by Salsberg and coauthors found a 1 percent annual growth in ACGME entry positions.

Note 3. As of 2013, programs wanting to participate in the NRMP had to offer all of their positions in the match (“all in” policy).  This led to a large increase between 2012 and 2013.  Some of the increase in 2014 may have also reflected existing programs deciding to participate in the match. The change in the NRMP administrative policies makes comparisons with NRMP data prior to 2013 inappropriate except for US MD seniors who have been required to participate in the NRMP for many years.