Prior training varies by residency program

U.S.-trained MDs, DOs, and international grads often train separately.

Although U.S.-trained allopathic medical school graduates (USMDs), osteopathic school graduates (DOs), and international medical graduates (IMGs) are all eligible for the same internal medicine residency slots, programs are highly segregated by graduate type, a recent study found.

To assess the proportion of USMDs, DOs, and IMGs (both U.S. citizens and noncitizens) in internal medicine residencies, researchers looked at 476 nonmilitary programs in the 2017 to 2018 year. As a proxy for program quality, they also looked at American Board of Internal Medicine certification exam pass rates for 388 accredited programs. USMD-dominated programs were defined as having more than 65% USMD residents, whereas DO- and IMG-dominated programs had fewer than 30% USMDs with more DOs or IMGs, respectively.

Overall, 25% of programs were USMD-dominated, 17% were DO-dominated, 42% were IMG-dominated, and 16% were truly integrated, according to results published online in December 2019 by the Journal of General Internal Medicine. USMD-dominated programs had significantly higher pass rates than the other categories. University hospitals were more likely to be USMD-dominated and were less likely to be IMG-dominated than community hospitals.

Lead author Tania M. Jenkins, PhD, an assistant professor of sociology at the University of North Carolina-Chapel Hill, and senior author Shalini T. Reddy, MD, MHPE, FACP, associate program director for the internal medicine residency program at John H. Stroger Jr. Hospital of Cook County in Chicago, spoke with ACP Hospitalist about the implications of these findings.

Q: What led you to study this issue?

A: Dr. Reddy: [Dr. Jenkins and I] connected because I've had a longstanding interest in medical education and, increasingly, the disparities that exist in training between international graduates and American graduates.

Dr. Jenkins: In the course of my research for my forthcoming book, “Doctors' Orders: The Making of Status Hierarchies in an Elite Profession,” I started to notice that residency programs seemed to be segregated on the basis of medical pedigree. So-called “IMG- or DO-friendly” programs are commonly cited on websites like, but when we turned to the literature, we couldn't find studies documenting the distribution or the impact of IMGs, DOs, and USMDs concentrating in different internal medicine programs nationwide. So we decided to look for the data and produce our own study.

Q: Were you expecting to uncover this segregation?

A: Dr. Reddy: Absolutely, because I've worked at four different institutions. I've worked at Parkland Hospital in Dallas, which is a county hospital. I did my training there, and . . . it was probably about 80% American and 20% international, and then I was at the University of Chicago, which was 100% American grads. And then I worked at Mercy Hospital [in Chicago], which was about 60% IMGs (either U.S. IMGs or foreign-born IMGs), and the remainder was DOs. And now I'm at Cook County, which is 80% international grads and 20% American grads. So I've been in all the different settings, and I've really seen that the international grads have to almost be better than American grads in order to get their foot in the door. . . . It's a little harder for me to speak broadly about DOs because I really have close exposure to [medical students from] one or two schools . . . and the quality of their training is outstanding. I don't really see a huge difference between those who are trained in osteopathic medicine versus allopathic medicine in terms of their skill set. What I know less about is the curriculum of different medical schools.

Q: What were your most interesting findings?

A: Dr. Jenkins: Perhaps the most interesting finding is that only 16% of internal medicine programs nationwide can be described as “integrated,” meaning that between 30% and 65% of the residents are USMDs; the rest have high concentrations of USMDs, IMGs, or DOs without mixing of the types of graduates. I think we expected patterns of segregation, but not to this extent.

We also found the difference in board pass rates to be troublesome. International graduates tend to perform just as well—if not better, in certain cases—than USMDs on USMLE [U.S. Medical Licensing Exam] Step 1 before residency. The fact that programs with large concentrations of IMGs tend to have lower board pass rates at the end of residency may suggest differences in training quality between USMD- and non-USMD-concentrated programs. Future research is needed to test this possibility and to rule out alternative explanations, such as unobserved differences between residents who enter USMD-concentrated programs versus non-USMD-concentrated programs.

Q: What are the potential reasons why USMDs, IMGs, and DOs concentrate in different types of hospitals?

A: Dr. Jenkins: It is unlikely that there is a single reason behind the segregation. . . . Certainly self-selection likely plays a role, with USMDs preferentially applying to higher-prestige university programs with large numbers of USMDs, and IMGs and DOs applying to so-called “friendly” programs, typically in community hospitals. However, we know from the literature that self-selection is not the whole story.

Other explanations could be that segregation may be based on merit. Could it be that USMDs are simply stronger candidates than non-USMDs? This is unlikely because we know that IMGs tend to perform at least just as well on the USMLEs as USMDs, making medical knowledge a less likely explanation for their exclusion from university-based programs. Another potential explanation could be that program directors have implicit or overt bias against non-USMDs, perhaps related to institutional pressures to increase program prestige or concerns about how programs with large numbers of IMGs or DOs are perceived. We lack large-n studies, however, systematically examining the criteria program directors use to select their housestaff, so more research is needed on how the residency selection process may be contributing to segregation.

Dr. Reddy: I think when you haven't worked with international grads, there is a perception that they don't have the equivalent level of skills as American grads. . . . The reality is that for international grads, it's so much harder to get into medical school because many countries have free medical education, and in order to even get in, you have to be the best of the best. And then, to come over to the United States, you have to be the top of the top. So most international grads have had to jump through significantly more hoops than American grads. But I think if you're a program director and you've never worked with international grads, you don't see that perspective. You don't see the quality of IMGs. I think the other thing that program directors are worried about is the perception, “Will prospective candidates think less of the program because they have a high concentration of international grads?” I think when you look at university programs, there's a lot of fear of how they'll be judged because of that.

Q: What are the potential implications of these findings?

A: Dr. Jenkins: Sociological studies tell us that segregation can form the basis for very powerful social differences in status between groups like USMDs and non-USMDs and can lead to unequal treatment within the profession. As such, the findings raise questions about whether the segregation of residents by medical school pedigree is in fact desirable, particularly given broader efforts to diversify the medical workforce. More integrated programs might benefit from having USMDs, DOs, and IMGs sharing their varied clinical experiences. The results also raise questions about whether all trainees are getting the same quality training and, eventually, whether this might matter for patient care.

Dr. Reddy: The reason it's a problem is because we don't have enough U.S. graduates applying into internal medicine to fill all of the spots we have in internal medicine. If international grads didn't step in to fill those spots, we would have a huge shortage of residents. . . . The reason that having segregation is problematic is I think it just carries forward stereotypes and perceptions that international grads are lower quality than American grads. And that is a real problem because, having worked with international grads, I will tell you that I would put any of my residents up against a person at a university program and they are as good and, most of the time, even better.

Q: What actions could residency programs take in response?

A: Dr. Jenkins: Insofar as segregation may be worsening status hierarchies between USMDs and non-USMDs, and given the benefits of a more integrated housestaff, residency programs may want to consider recruiting more candidates from diverse backgrounds. To reduce bias, program directors could review applications blindly (without knowing the applicants' medical school of origin) or aim to titrate their rank lists to reflect the broader composition of the residency workforce, with USMDs comprising roughly 60% of all matched residents nationwide. We may also want to take a closer look at concentrated programs with lower board pass rates. At present there are insufficient numbers of USMDs to fill all the available [internal medicine] residency spots; IMGs and DOs are crucial to allowing us to appropriately staff our teaching hospitals. Ultimately, if we continue relying on non-USMDs to fill gaps in health care, at a minimum, we need to make sure programs have the resources to train them adequately.

Dr. Reddy: From my end, I'm very happy with the residents that we have. We have incredible residents, and I think that university programs would probably feel the same way: “Why should we take international grads [up] to an unknown quantity? There are other issues that we have to worry about. We're perfectly happy with who we have.” So I think there's satisfaction with it. I would encourage university programs, though, to at least consider what they're missing. It works well for me because if they ignore the top of the top, then we get them. I have the three smartest people from Nepal in my residency program, and I'm very grateful for that. I don't particularly want a university program to start taking all my outstanding residents, but I also feel like they're missing out.