The COVID-19 pandemic has turned medical education upside-down. From exclusively virtual pre-clerkship courses, to delayed clerkships, to canceled graduation proceedings, there has been massive disruption. Now, after weeks of speculation, we’ve learned how the pandemic will be disrupting the 2020-2021 residency application cycle. The Association of American Medical Colleges (AAMC) is now recommending that all residency interviews be conducted virtually. While this a wise recommendation, out of concern for public safety, I fear it will exacerbate long-standing problems in the residency application process and will have dire consequences for many applicants.
For over 20 years, there has been hyperinflation in the number of residency applications submitted per applicant. Compared to 2001 applicants, 2020 applicants applied to 3.14x the number of programs in anesthesiology, 2.31x in emergency medicine, 2.56x in general surgery, 2.71x in internal medicine, 2.62x in OB/GYN, 1.76x in orthopedic surgery, and 4.02x in psychiatry (Electronic Residency Application Service (ERAS) data for U.S. and Canadian graduates only). While resources like the AAMC’s Apply Smart for Residency and the University of Texas Southwestern’s new Texas STAR database provide valuable insights to applicants about the correct number of applications to submit, little headway has been made yet in helping curb the hyperinflation.
The downstream consequence of this is that residency interviews have become a ‘tragedy of the commons,’ in which every applicant attempts to gain as much as possible from the common resource—interviews. Above-average applicants, who do not need as many interviews to match, are able to crowd out (by no fault of their own) below-average applicants by applying to more programs and accepting more interview offers than are necessary. This is especially consequential when above-average applicants accept interview offers from below-average programs—‘safe schools’—just so they can have more confidence in matching, when they could leave those interviews open for other applicants who need them.
Figure 1: Average number of residency applications and interviews per matched U.S. allopathic medical school graduate
Data obtained from ERAS reports and National Resident Matching Program (NRMP) Charting Outcomes in the Match publications. Number of contiguous ranks for matched U.S. seniors was used as a surrogate for number of interviews, as the data on interview offers and accepted interviews is not publicly available.
Thankfully, there has always been one factor limiting the damage of hyperinflation and interview hoarding—time. An above-average applicant could have dozens of interview invitations, but time constraints (not to mention financial concerns) limit the number of invitations they can accept. One simply cannot interview in California in the afternoon, and make it to Maine by evening for a pre-interview dinner. But now, with interviews going virtual, I fear they will be able to accept all, or nearly all, of the interview invitations offered them.
While applicants have this newfound capacity to interview at more programs, programs do not have the ability to proportionally increase their interview invitations, given the time required of faculty interviewers. Consequently, the invitations above-average applicants would have typically declined will not get passed on to below-average applicants this year. As such, I anticipate a greater disparity in the length of rank order lists and match outcomes between above- and below-average applicants if immediate action is not taken.
Previous solutions to application hyperinflation have been proposed, such as limiting the number of programs an applicant can apply to. However, these have seemingly been met with apathy and lack of urgency, presumably because match rates have not significantly changed in response to hyperinflation. But this year, I fear will be different.
I can find no other option to effectively avert the coming disaster than for the AAMC to recommend a limit of fifteen interviews, per applicant per specialty. Limiting the number of interviews an applicant takes will help level the playing field between applicants of different calibers. This would prevent, say, an above-average applicant from interviewing at twenty-five programs, while a below-average applicant only interviews at five.
Based upon the NRMP’s 2018 Charting Outcomes in the Match, fifteen contiguous ranks provide a >95% probability of matching in every specialty, except for neurological and vascular surgery—small fields in which greater interview parity will still likely result in better match rates. Additionally, by recommending an interview limit, I hope there will also be a consequent decrease in the number of applications per applicant, thus curbing the ongoing hyperinflation.
The AAMC could further encourage programs to schedule their interviews through ERAS rather than third-party companies, in order to increase compliance with the interview limit recommendation. In response, the NRMP could also enforce a rank order list limit of fifteen, to further promote adherence.
While my proposed limit of fifteen interviews may not be the appropriate number, the organization that has the data to determine a proper limit also has the power to publish a nationally obeyed recommendation—the AAMC. While it has taken the appropriate step to encourage that all interviews take place virtually this year, its leadership needs to take additional steps to avoid a disastrous 2021 match.
Jordan Hughes is an emergency medicine resident.
Image credit: Shutterstock.com