Breaking residency placement fever

 By Francis Deng* and Skeptical Scalpel

A recent opinion piece entitled “Residency Placement Fever” in the journal Academic Medicine by Gruppuso and Adashi noted a recent intensification in the volume of residency applications submitted and interviews offered/attended per applicant.

For keen observers of the Match process, this trend is neither a secret nor a surprise. The Electronic Residency Application Service (ERAS) has seen an increase in applications filed per US medical school graduate from an average of 30.3 in 2005 to 45.7 in 2015.

Inflation of residency applications is getting out of control these days. Time for a limit in ERA$ pic.twitter.com/AsRn8c3fgd

— Francis Deng, MD (@francisdeng) October 31, 2016

The biannual surveys of applicants in the Main Residency Match conducted by the National Resident Matching Program (NRMP) show a recent increase in interviews for many specialties.

Medical students are going to more residency interviews than ever before. If only travel points paid down school debt! pic.twitter.com/NYEbeMCFLJ

— Francis Deng, MD (@francisdeng) December 4, 2016

Gruppuso and Adashi claim it remains unclear “whether or not the number of GME program interviews per applicant is predictive of a successful match.” The NRMP data, however, are stratified by match outcome. Generally, unmatched US senior applicants apply to more programs but have fewer interviews compared to their matched peers.

Why is this happening?

The authors pin the blame for the spread of residency placement fever on students themselves, who “appear to be keeping their own counsel against the advice of medical school advisers and mentors advocating moderation.” They point the finger at “the perception that the attainment of a match has become ever more difficult.” But this febrile illness does not originate endogenously with each new batch of applicants. Advisors and institutions spread the bug and fan the flames year to year.

For instance, as part of a lobbying effort to increase federal funding for graduate medical education (GME), the AMA raised the alarm of unmatched US medical students as evidence for a residency bottleneck. The notion of a GME squeeze, however, has been debunked as far as US medical graduates are concerned, for whom there will continue to be more positions available than applicants seeking them.

The AAMC Organization of Student Representatives counsels conservatively, “it is wise to apply to more programs than you think you will need.” The AAMC Roadmap to Residency generally recommends 30-40 applications per student. Advisors make minimum recommendations based on average numbers from prior years, thereby perpetuating inflation.

The longstanding advice of the NRMP has been that applicants are more likely to be successful if they are able to rank more programs. The length of rank order lists among US seniors increased from an average of 8.1 in 2005 to 11.5 in 2015. Nevertheless, the match rate for US seniors remains steady at or above 94% and the vast majority (85%) of those US seniors who matched in 2015 did so to one of their top four choices.

Students are not the ones to blame; they are the victims of a system that has spiraled out of control. Applicants find themselves in a prisoner’s dilemma, where they must apply to more programs to maintain the same chance of success because everyone else is submitting more applications and adding noise to the process. This arms race is futile, as yearly variations in application number are not associated with specialty-specific match rates. The presumption is not simply that “more is better,” but that “more is necessary” for the individual even though the collective has already reached a point of diminishing returns.

What is the impact?

The downsides of this trend toward increasing numbers of applications are many. The ERAS fee structure makes applying to 15 orthopedic programs cost $159, but applying to 73 programs would result in fees of $1497. The more interviews an applicant attends, the more travel, lodging, and food expenses mount. Finally, interviews take time away from the fourth year of study. In the status quo, the only brake against over-applying and over-interviewing is the financial and educational costs that applicants bear. On the program side, faculty have to spend inordinate amounts of time to review applications and interview.

How do we treat the fever?

The authors offer two proposals to cut down interview costs. First, programs can conduct brief interviews online and only offer in-person interviews to those who pass that hurdle. That has potential, but having interviewed prospective residents for more than 25 years, the senior author of this post is not so sure he could effectively evaluate applicants that way. It is hard enough to do so in person.

Second, there can be a cap on the number of interviews per student. Interview offers are not centrally coordinated, however, making a cap unenforceable. A more feasible measure would be to limit the number of applications submitted through ERAS, as recently proposed by various researchers and educators. This tackles the root of the problem, as application volume has grown much more rapidly than interview volume.

In a recent NRMP survey, medical school advisors seemed to be reluctant to impose a limit on applications, believing there may be negative consequences for students, while residency programs were more receptive to a limit. At the same time, schools expressed the need for applicants to get program-specific information to adequately assess their competitiveness, such as USMLE Step 1 and 2 score cutoffs used in selection and average statistics from previously interviewed and matched applicants. The NRMP provides such statistics, based on applicant-reported and school-verified data, to individual programs confidentially. Many programs, unfortunately, are reluctant to share these reports and their selection methods.

Insecurity erects barriers to transparency. Years ago, when surgical programs self-reported data about the characteristics of their residents for a book called “So You Want To Be a Surgeon,” a.k.a. “The Little Red Book”, the program data was almost laughably inaccurate. About 25% of programs claimed that some or most of their residents were AOA graduates. Many programs claimed that more than 75% of their residents were graduates of US schools when it was well known that the actual percentage was the inverse of that number or less.

Mutually assured defervescence

A compromise would involve an application service that programs can opt into only if they share detailed program-specific data, and in exchange, students may only apply to a limited number of programs in this group. With a restricted residency application scheme, programs could focus their energies on suitably competitive applicants who have the most interest rather than sift through all comers who just checked another box. As all students applying to these programs would be submitting fewer applications, individuals would maintain the same chance of success as before.

To move forward in restoring sanity in the residency placement process, there needs to be greater awareness in the medical education community of the extant data describing “application fever” and various stakeholders’ feelings toward the proposed solutions.

*Francis Deng is a resident at Brigham and Women’s Hospital and former member of the board of directors of the NRMP. He tweets @francisdeng.