American College Of Surgeons - Inspiring Quality: Highest Standards, Better Outcomes

Did You Know? The History of “The Match”

Dr. HendrenAs we approach the mid portion of the academic year, the fourth-year medical students among us are making career choices and refining their list of preferences for postgraduate training. Once they have determined their preferences, they submit them in rank order to the National Intern and Residents Matching Program, which, with modification to accommodate changes in computer technology and to accommodate couples matches, has been based on a durable algorithm relatively unchanged since 1952. “The Match” seeks to pair each applicant with a postgraduate training program by systematically identifying the applicants’ preferences and simultaneously identifying available training positions. As the program is run, the postgraduate training positions fill with the most talented trainees identifying and matching with their highest preferences first; as these positions fill, secondary choices of the applicant are considered until all positions are filled.

Prior to 1952, residents and hospitals alike were participants in a scramble—a decentralized process offering no assurance that residents could either make an informed decision at a reasonable point in their training or secure the best possible position for themselves.

Well-intentioned individuals representing hospitals, testing services, and the health care industry, funded by a $100,000 grant from the American Association of Medical Colleges, convened in the late 1940s and early 1950s to remediate this situation and to attempt a computer-assisted match, a now-familiar tool that reconciles applicants’ and hospitals’ preference, granting one acceptance per applicant on a date in mid-March. When a final product of their deliberations was produced, the dean of Harvard Medical School, George Packer Berry, MD, presented the match algorithm to the Harvard third-year medical students on October 15, 1951. His counterparts around the country did the same.

Among the Harvard students in Dr. Berry’s audience was a recent transfer from Dartmouth Medical School, who put his subsequent training at risk by pointing out that the program as presented was flawed, as it did not assure the student the best program attainable.

Computers at that time were still paper-based, and programs were run with a punch card system requiring manual feed of IBM cards. In the case of the newly minted match program, all was well on the first pass if hospital and trainee both ranked each other first. After that, however, the student could potentially have to wait until a third round, or perhaps a fourth, to get an opportunity to be considered for his next choice (but by then, of course, that hospital may have already filled with alternatives that had ranked that institution first).

The student in question deliberately pointed out this shortcoming and suggested an alternative. The dean was unimpressed, but his assistant dean, Reginald Fitz, MD (son of the Massachusetts General Hospital pathologist who in 1886 described a series of patients with appendicitis), was persuaded by the student’s argument.

Emboldened by this course of events, that student organized a small group of students who met and thought that it would be appropriate to convene a larger group from around the nation. The student borrowed $3,000 from his father and secured the services of a secretary and a printing company. The class president of every medical school in the country was invited to a meeting held on October 21, 1951, at Columbia University College of Physicians and Surgeons, a location chosen to maximize attendance. Also invited were some of the architects of the plan, including not only Dr. Berry but George Stalnaker, head of the Princeton Educational Testing Bureau; Dr. William Crosby, president of Johns Hopkins Hospital; and Dr. Joseph Mullen, Dean of the University of Chicago. Engineers and representatives from IBM were also present and served as technical support. Initially, the senior people in the room defended the match algorithm as proposed. The students, however, brought Mr. Louis Servesio from the accounting department at Mass General, who agreed with the flaws as identified by the student group.

The secretary hired by the students memorialized the discussion and mailed ballots offering a choice of computer programs to close to 5,000 students nationwide. With an 83 percent response rate, the modified plan (by then known as the Boston Pool Plan) won handily over the original. However, on November 9, the stakeholders in the original plan, together with representatives of all hospital associations, met in Chicago, IL, and concluded that November was too late to change a plan whose results would be reported in mid-March. The students stood firm, and the revised plan that they had proposed survived as the match. Skeptics in the group, including Dr. Berry, were tenacious and strident in the opposition, convinced that it was the new program that was flawed. Dr. Berry promised the student that if the program ran unsuccessfully, the student would not graduate from Harvard Medical School.

Fortunately, the plan was sound, and the program ran flawlessly. Ironically, the dean was congratulated for his open mindedness and his willingness to allow student participation in a key process that determined their fate. The computer program itself has survived over the years with very little change and has made the transition from punch cards to paperless. The student in question went on to surgical residency at the Massachusetts General Hospital and Boston Children’s Hospital. He became the distinguished Robert E. Gross Professor of Surgery at Harvard, the chief of surgery emeritus at Boston Children’s Hospital, and honorary surgeon at Massachusetts General Hospital. The student, W. Hardy Hendren III, MD, FACS, is the recipient of more than two dozen major surgical honors and has had an endowed chair in surgery established in his name at Harvard. Three endowed fellowships have also been established in his name by grateful patients.