Monday, March 28, 2011

Problems with the Current NRMP Match

Fifty years ago the medical community was faced with a worsening problem. In many specialties, the competition for the best and brightest medical students was at an all time high. Medical students were forced to commit to residency programs while still years from graduating, or risk doing residency at undesirable programs. Hospitals were forced to play the game or risk losing out on top applicants. Then the National Resident Matching Program (NRMP) arrived on the scene.

Established in 1952, the NRMP created a new process for residency programs to find medical students. Now programs and students would "rank" each other, then submit their respective lists to the NRMP (first on paper and now electronically. A computer algorithm match medical students to programs, and order was restored to the world. Or at least that is what they want you to believe. From their website:
"The National Resident Matching Program (NRMP) is a private, not-for-profit corporation established in 1952 to provide a uniform date of appointment to positions in graduate medical education (GME). It is governed by its Board of Directors. Five medical/medical education organizations, one program director organization, and three medical student organizations nominate candidates for election to the Board: the American Board of Medical Specialties (ABMS), the American Medical Association (AMA), the Association of American Medical Colleges (AAMC), the American Hospital Association (AHA), the Council of Medical Specialty Societies (CMSS), the Organization of Program Director Associations (OPDA), the AAMC Organization of Student Representatives, the American Medical Student Association (AMSA), and the AMA Medical Student Section. The Board also selects one program director, three resident physicians, and one public member from at-large nominations. Each year, the NRMP conducts a Main Residency Match that is designed to optimize the rank-ordered choices of applicants and program directors. In the third week of March, the results of the Match are announced.
The NRMP is not an application processing service; rather, it provides an impartial venue for matching applicants' and programs' preferences for each other consistently. Each year, approximately 16,000 U.S. medical school students participate in the Main Residency Match. In addition, another 20,000 "independent" applicants compete for the approximately 25,000 available residency positions. Independent applicants include former graduates of U.S. medical schools, U.S. osteopathic students and graduates, Canadian students and graduates, and students and graduates of international medical schools."
Unfortunately, things are not quite as perfect as they would like you to believe. First, the match is expensive. Both medical students and residency programs pay a significant amount of money each year to the NRMP for them to do three things: ensure that medical students are eligible to participate, maintain a website allowing for list entry, and running the lists through a basic computer algorithm. For this they collect roughly seven million dollars per year.

But more than just a financially, there are several problems that arise when one considers the actual mechanics of the match. If you are not familiar with the process, it can be found here. The most simple way of thinking about it is that medical students are guaranteed to match at their top choice if the program ranks them within as many open spots as they will have. For instance if the University of Chicago has 5 spots in their orthopedic residency, then any medical student in their top 5 ranks--who ranks them first--will match there.

As you can see from the description, the match really favors residency programs. No matter how much they try to tell medical students that they have the advantage because their list is "looked at" first by the computer, program directors really have the final say. I could want to get into a plastic surgery residency at the Mayo clinic, but unless I am ranked in their top few spots, it is not going to happen. In less competitive specialties students lists are emphasized to a greater extent, but the power still resides with the programs. I do not necessarily have a problem with programs being favored, my problem is more with the deception that medical students have the power to determine their destination.

A related problem is that "the Match" does not actually create the best possible matches. Using the same line of thought as above, a program may match a medical student that ranked them 15th, while not matching a medical student who ranked them first. Although the program may be happy that they got someone from the top of their list, in actuality they are getting a future resident that didn't really want to be there. This seems like a prescription for losing residents who drop out or refuse to sign a contract.

Those that do match did so through a program that is unlike anything else in society. When does one apply for jobs and expect to wait 4 months to find out where they matched. "Well I ranked Subway over Burger King but I'd really be happy going anywhere in my top three..." Intentionally or not, the match has created a system that restricts concrete thinking about where one would best fit in. And because of how closely guarded the data is, it is difficult to determine just how poorly the match performs.

Which brings up another problem. The match is trying to do all things for all people. As mentioned before, the match seems to work best for specialties that are not very competitive such as family medicine. When a program has to rank 300 applicants to fill 20 spots, there is a greater emphasis on where a medical student would like to end up. As described above, this works less well for competitive specialties. Because of this and the lack of information provided by the NRMP, several specialties have formed their own matches.

Those that do not match must either wait another year or take part in something called "the Scramble." During this time, medical students that did not match contact programs that did not fill all of their spots. It is impossible to describe the chaos of trying to contact the a dozen programs that have hundreds of medical students calling at the same time. Because of this many programs that have spots open are filled through back room deals via channels of communication within the specialty.

Having just gone through the match process I can say unequivocally that the system is flawed and in dire need of modernizing. It is expensive and stressful and often does perform its one task well, to best match medical students to programs.

As always, feel free to put any thoughts on the Match in the comments.

3 comments:

ZH said...

What are your thoughts on applying to programs that don't participate in the match? There are number of programs on FRIEDA web sites. I am assuming they need resident too.

Susan said...

You forgot to (or perhaps, intentionally did not) mention a few salient points. The student is favored because they are protected by being able to freely rank the programs they truly prefer first (regardless of how competitive) while not compromising their safeties that they listed lower. The match allows this by solidifying your spot on the program director's rank list. NO ONE ranked below you on any list can get your spot at that program unless you match at somewhere you ranked higher. So then you are free to rank whatever places you truly prefer like perhaps Harvard and Yale even though you have a slim chance of matching there because your spots at the other places are safe. That is why the applicant is favored.

Anonymous said...

1. Ditto above that the student is preferred insofar as it doesn't hurt you to rank your most preferred programs first.
2. "And because of how closely guarded the data is, it is difficult to determine just how poorly the match performs." The NRMP releases this data. Over the past decade, about half of applicants get their first choice. http://b83c73bcf0e7ca356c80-e8560f466940e4ec38ed51af32994bc6.r6.cf1.rackcdn.com/wp-content/uploads/2013/08/resultsanddata2013.pdf