Since publication of the Flexner Report1 in 1910, U.S. and Canadian medical schools have almost exclusively followed a four-year, postbaccalaureate graduate degree model. On earning the MD degree, physicians require additional years of clinical training for board and subspecialty certification. Those physicians who wish to pursue medical research as well as clinical work spend many more years in research training to acquire the requisite skills. Medical school deans and the Association of American Medical Colleges (AAMC) have raised concerns about the effects of prolonged clinical training and encumbered debt on the ability to produce physicians who also will lead cutting-edge research and innovation in medicine.2,3
The effect of the duration of medical training on innovation is underscored by the fact that, overall, physicians are less likely to receive extramural National Institutes of Health (NIH) funding than are PhD scientists, and those physicians who are successful obtain funding at a later median age than do PhDs (43.4 years versus 41.7 years, respectively).4 And compounding the opportunity cost of years in training, the practice of medicine is becoming more challenging as remuneration is declining. It is time to consider altering medical school curricula to reduce training time and debt burden or to attract trained scientists into the practice of medicine—all without sacrificing physician quality.
We believe this discourse can be informed by the experience of the University of Miami School of Medicine's PhD-to-MD Program, which was in place for 18 years during the 1970s and 1980s. With this program, the school pioneered a radically different approach to medical education that was designed to enable PhD scientists to obtain an MD degree in two years. In 2008–2009, we conducted a retrospective review of program graduates to determine student outcomes and to inform medical educators about alternative approaches to developing physician–scientists. Here we report the available long-term outcomes of the unique program, whose graduates were at least 20 years from the awarding of their medical degrees at the time of our study. The institutional review board at the University of Miami approved this study.
The PhD-to-MD Program
In response to reports of a projected national physician shortage, the University of Miami School of Medicine in 1971 instituted a pilot program to provide an accelerated medical school education to individuals with a PhD in the biological, chemical, or physical sciences.5 The accelerated curriculum, known as the PhD-to-MD Program, conferred a medical degree in 24 months without any prerequisite courses. The intent was to take advantage of the graduate training that PhD scientists had already received because the course work for such programs overlapped heavily with the standard medical school curriculum. Quickly, however, the program was extended to include scientists with PhDs in physics, engineering, or mathematics. As reported elsewhere,5 the matriculants with nonbiological degrees mastered the accelerated curriculum material and performed very well on certification exams. All accelerated program matriculants paid a higher annual tuition rate than did their four-year counterparts, so their total tuition was equivalent to that paid by the traditional medical students.
Shortly after the program's inception, medical school enrollment expanded at existing schools and new medical schools opened to help address the predicted U.S. physician shortage.6 The medical school decided to continue the PhD-to-MD Program because it was attracting an impressive group of highly qualified and accomplished scientists into medicine.7
State medical licensing boards, however, raised concerns regarding the number of months of medical school enrollment.6 Statutory requirements, initially put forth by the California Medical Board, required that medical school be at least three years in duration. Such requirements resulted in part from concerns over licensing physician graduates from foreign medical schools that offered similarly shortened curricula. Although the medical school could have argued that the total amount of training the PhD-to-MD students received to obtain the MD degree far exceeded a standard four-year curriculum, it voluntarily terminated the program at the request of the Liaison Committee for Medical Education in 1987. The program graduated its last class in 1989.
During the program's 18 years, it accepted approximately 10% of applicants solely on the basis of their academic and professional records of accomplishment; there were no other prerequisites. Almost all of the 537 matriculants had multiple prior publications in their respective disciplines. The majority (n = 283, 53%) of students had PhDs in biochemistry or the biological sciences, followed by chemistry (n = 84, 16%), physics (n = 79, 15%), engineering (n = 74, 14%), mathematics (n = 14, 3%), and oceanography (n = 3, 1%). The program conferred 525 doctorates of medicine. Of these graduates, 508 completed their MD within the two-year period, and 17 required longer times to graduation (12 students had personal reasons, and 5 had academic reasons). Ten students withdrew from the program, and two were dismissed.
Course rotations and examinations
The condensed timeline of the PhD-to-MD Program necessitated a preclinical curriculum separate from the traditional medical school's (Table 1). The program's students began medical school in late June and attended nearly 40 hours of lectures per week. They devoted their first six months in the program to 935 hours of basic science instruction and the following two-and-a-half months to 316 hours of courses on the introduction to medicine and mechanisms of disease. Instructors generally scheduled exams on Monday mornings to allow the students maximum time to review course work over the weekend.
Following the didactic portion of the curriculum, students undertook one year of clinical rotations identical to those taken by third-year medical students in the regular four-year curriculum, including rotations in medicine, surgery, pediatrics, obstetrics–gynecology, psychiatry, neurology, and anesthesiology. Students finished the program by completing approximately three months of elective rotations, generally including a subinternship/externship. During the two-year program, students had a total of six weeks of vacation: two weeks before the start of clinical rotations and a two-week period during each year's winter holidays.
PhD-to-MD Program students' test scores both on preadmission and licensure examinations were considerably better than both those of the medical school students taking the standard four-year curriculum and the national averages over a similar period (1973–1989).8 Many of the program's graduates moved on to highly competitive residency programs and other medically relevant training programs.8
Evaluating Career Outcomes
To generate as complete a list of PhD-to-MD Program graduates as possible from the largely precomputer era of the PhD-to-MD Program, we collated student graduation lists from program files and graduation logs. The university registrar subsequently cross-referenced our lists with the medical school's list of known graduates. We then cross-referenced our lists of medical graduates by unique identifiers with AAMC lists of faculty at U.S. academic medical centers. We identified 487 (96%) of the 508 graduates who completed the program in two years.
We obtained additional information for each graduate by conducting public searches by name on Google and by query of the Social Security Death Index. We identified graduates' publications through searches of the ISI Web of Knowledge and the Scopus citation and abstract databases, and we identified their grants by query of the NIH's Computer Retrieval of Information on Scientific Projects database. Patents were identified through searches of the U.S. Patent and Trademark Office and Scopus databases. To verify our attribution process, we cross-referenced publications and grants attributed to particular graduates with the graduates' identified institutional histories. The numbers of publications, grants, and patents we report in this article likely underestimate the total for the group because we did not include in the totals in Table 2 those written or received by graduates with common names whom we could not distinguish from others in database searches.
Of the 487 graduates we identified, 5 died during residency training and 4 others have died since. In 2009, PhD-to-MD Program graduates were an average of 28 years (range: 20–36 years) from the awarding of their MD and were an average of 60.9 years old (range: 43–74 years old). Their areas of medical specialization following graduation are shown in Figure 1. The most popular was internal medicine (n = 154, 32%), followed by anesthesiology (n = 67, 14%) and radiology (n = 48, 10%). As did others in their reports on combined MD–PhD programs, we found that few students chose surgery or surgical subspecialties.9
On completion of training, 67% (n = 323) of the 482 surviving students went into academic, government, or industry positions; most of these entered university positions (Figure 2). The other 33% (n = 159) entered private practice. This distribution is similar to the distributions reported for many combined MD–PhD programs.10 Overall, by 2008–2009, 29.6% (n = 85) of those entering university positions had attained the rank of full professor and 48.9% (n = 139) associate professor. Among the graduates were 11 university directors or division heads, 14 chairs of academic departments, 2 directors of university cancer centers, 2 deans of medical schools (1 in the United States and 1 in the United Kingdom), 3 directors of institutes at the NIH, and 1 astronaut.
We also used available measures to determine academic success and productivity (Table 2). Of the 482 two-year medical school graduates, 28.2% (n = 136) were the principal investigators on post-MD, NIH-sponsored research programs. We noted Medline or Scopus publications after receipt of the MD for 56% (n = 270) of the graduates. Most papers and all grants were attributed to those graduates in academic positions.
The long-term career outcomes of the PhD-to-MD Program graduates indicate that the program's accelerated medical training of scientists with a record of successful, rigorous research produced a highly successful group of academic physicians.
Lessons for Future Programs
Producing innovative, research-oriented physicians who are also competent in the practice of clinical medicine continues to be a challenge.2,11,12 The time and expense, both direct (tuition) and indirect (opportunity cost and deferred compensation), discourage many students from pursuing careers combining research and academic medicine.
Reducing time in training may be one solution to the problem of attracting trainees. Various researchers and oversight organizations have suggested approaches to streamline postgraduate training during residency and fellowship.13 Some of these pathways, including those suggested for surgery, would abandon broad training and concentrate on a standardized set of diseases treated by a specific subspecialty selected by the individual. Shortening of specialty training in internal medicine has also been proposed and implemented.14 This fast tracking means that trainees spend less time in their internal medicine residencies and have less general medicine experience prior to subspecialization.
The University of Miami School of Medicine's PhD-to-MD Program experience suggests that early compression of the curriculum in medical school may be an attractive and effective alternative means of reducing training time. The success of the program graduates on the clinical licensure examinations indicates that they were able to master the medical curriculum in half the traditionally allotted time, even though nearly half of the matriculants held PhDs in fields outside the biological sciences. Adopting such early compression in today's MD–PhD programs would allow more time for subsequent specialty and subspecialty training.
Our analysis of the program graduates' outcomes indicates that compressed medical school training is sufficient both for graduates to achieve successful medical careers in their respective fields and for a significant percentage to become involved in academic research. About one-third of the program graduates entered private practice. However, approximately 60% to 70% of graduates entered academic (university or government) careers, which is consistent with the results reported by conventional combined MD–PhD programs.15 Furthermore, the program's graduates became high-achieving physician–scientists in all three realms: Many who eschewed private practice rose to the highest levels of academia (39.4%), industry (67%), or government (40%). It is unclear, though, whether the PhD-to-MD students were less, similarly, or more productive than were graduates of conventional combined MD–PhD programs; as of the time of this writing, we have not found any reports of similar long-term follow-up of combined MD–PhD programs.
It is important to note that, unlike many combined MD–PhD programs, the PhD-to-MD Program required significant financial investment from its students in the form of medical school tuition approximately equivalent to that paid by students in the regular four-year program. It remains unknown whether more PhD-to-MD graduates would have pursued academic careers if they had been provided tuition waivers and stipends similar to those given to their combined MD–PhD program counterparts.
These results also raise questions about the necessary duration of medical training within combined MD–PhD programs, which typically consist of a four-year medical curriculum with an additional PhD component. Other than the emphasis placed on integration of these degrees, there is little change in such programs from the standard medical curriculum, save a scheduled break for completion of the PhD and some transfer of research credits from the graduate program toward the medical degree. Furthermore, a substantial degree of course duplication or overlap often exists between the PhD and MD curricula. The outcomes for the two-year PhD-to-MD Program graduates suggest that current MD–PhD programs might be able to accelerate the medical portion of the combined degree. Doing so would allow students to spend more time on focused research or permit an overall shortening of medical training.
If the PhD-to-MD Program were to be restarted to once again introduce proven innovative researchers into medicine, some alterations of the program might be considered. First, financial support, possibly from a U.S. government source, should be provided. All of the PhD-to-MD Program graduates entered medicine at an older age (median: 33 years) and had to pay for their medical education while supporting themselves and, in many instances, their families. It seems likely that a larger number of qualified students would consider such a pathway if financial support were available. Furthermore, the financial debt that the program graduates encumbered likely prevented a number of them, as is well documented for other medical graduates,16 from pursuing longer training or an academic career. Second, the program selected students without considering whether the candidates wished to pursue academic careers. Perhaps more directed selection for those with an interest in academic medicine, combined with financial support, would further increase the number of physician–scientists produced.
In summary, the PhD-to-MD Program produced a large number of academically accomplished physician–scientists. Given the outcomes, medical educators should consider creating similar programs for PhD scientists and mathematicians at their institutions. Such programs could tap into the existing reservoir of well-trained scientists with proven research abilities, many of whom might hear a call to clinical medicine later in life. This program turned away 90% of its applicants, which suggests that a substantial pool of interested scientists exists. Providing stipends and tuition support predicated on pursuit of academic medicine careers would likely further increase the number and quality of interested applicants and might present a direct route toward improving the training and retention of physician–scientists in the United States. Finally, as our results suggest, shortening the medical education portion of combined MD–PhD programs would speed the completion of the degree without sacrificing achievement on objective measures of knowledge.
The institutional review board at the University of Miami approved this study.
1 Flexner A. Medical Education in the United States and Canada: A Report to the Carnegie Foundation for the Advancement of Teaching. Bulletin No. 4. Boston, Mass: Updyke; 1910.
2 Dickler HB, Korn D, Gabbe SG. Promoting translational and clinical science: The critical role of medical schools and teaching hospitals. PLoS Med. 2006;3:e378.
3 Heinig SJ, Krakower JY, Dickler HB, Korn D. Sustaining the engine of U.S. biomedical discovery. N Engl J Med. 2007;357:1042–1047.
4 Harris S. Accelerated M.D. programs provide benefits, but still face skepicism. AAMC Reporter. January 2009.
5 Harrington WJ, Whelan WJ, Fogel BJ, Papper EM. Alleviating the shortage of physicians. Science. 1971;172:1109–1111.
6 Awad WM. The P.D-to-M.D. Program of the University of Miami. In: Swartz HM, Gottheil DL, eds. The Education of Physician–Scholars. Rockville, Md: Betz Publishing; 1993:179–184.
7 Muench KH, Awad WM Jr, Whelan WJ. The experimental M.D. program for Ph.D.s: An evaluation at the three-year mark. J Med Educ. 1974;49:649–659.
8 Awad WM Jr, Harrington WJ, Papper EM. The Ph.D. to M.D. program: The seven-year mark. N Engl J Med. 1979;301:863–867.
9 Andriole DA, Whelan AJ, Jeffe DB. Characteristics and career intentions of the emerging MD/PhD workforce. JAMA. 2008;300:1165–1173.
10 Ley TJ, Rosenberg LE. The physician–scientist career pipeline in 2005: Build it, and they will come. JAMA. 2005;294:1343–1351.
11 Mallon WT, Korn D. Education. Bonus pay for research faculty. Science. 2004;303:476–477.
12 Fang D, Meyer RE. Effect of two Howard Hughes Medical Institute research training programs for medical students on the likelihood of pursuing research careers. Acad Med. 2003;78:1271–1280.
13 Fitzgibbons JP, Bordley DR, Berkowitz LR, Miller BW, Henderson MC. Redesigning residency education in internal medicine: A position paper from the Association of Program Directors in Internal Medicine. Ann Intern Med. 2006;144:920–926.
14 Debas HT, Bass BL, Brennan MF, et al. American Surgical Association Blue Ribbon Committee Report on Surgical Education: 2004. Ann Surg. 2005;241:1–8.
15 Bradford WD, Anthony D, Chu CT, Pizzo SV. Career characteristics of graduates of a medical scientist training program, 1970–1990. Acad Med. 1996;71:484–487.
16 Kerr JR, Brown JJ. Costs of a medical education: Comparison with graduate education in law and business. J Am Coll Radiol. 2006;3:122–130.