Physician scientists represent an essential national resource in the pursuit of clinically relevant basic and translational research discoveries. The recognition that this small group of physicians was an “endangered species” has sparked the establishment of a number of new programs by the National Institutes of Health (NIH) and private foundations to ensure that the pipeline of physician scientists who are committed to this career path will be maintained. A recent analysis indicates that these efforts are achieving some success.1
Although many of the initiatives are designed to encourage MDs to enter this pathway (such as the loan repayment programs of the NIH), the bulk of physician scientists continue to come from the medical scientist training programs (MSTPs) that were initially developed by the National Institutes of General Medical Sciences in 1964 and that continue to be supported by the NIH, and from other MD–PhD joint degree programs that are not financially supported by the NIH.2 Currently, there are 43 MSTP programs and several dozen other MD–PhD programs in the United States that are graduating individuals who are well trained in the elemental aspects of clinical medicine and biomedical research.3 In addition, many medical schools offer advanced research training to medical students that is outside the formal MSTP framework, such as MD–MS and MD–MPH programs.4
The graduates of those various programs require several years of additional clinical training to become competent to practice medicine in a particular subspecialty. Furthermore, those individuals almost always need postdoctorate research training, in part because they will have been away from the laboratory for five to six years while pursuing their clinical training. The postdoctorate training also allows them to enter new research areas of interest based on their clinical experiences.
This additional training follows an already prolonged joint degree program. As a result, individuals are usually in their mid- to late 30s before they are able to assume an independent faculty position. The prolonged training requirements and the appeal of lucrative alternative career options for physicians with clinical expertise create a climate in which some graduates consider abandoning their initial goal of becoming academic physicians to pursue a career in the private practice of medicine.
In reviewing the entire training path taken by individuals who enter MD–PhD programs, we noted that professional development after graduation from medical school was largely amorphous and unstructured. Because the graduates face a number of difficult challenges in their career development, we felt that a structured program tailored to the specific needs of these individuals could serve as a valuable adjunct to their training. Consequently, the physician scientist training program (PSTP) at Washington University School of Medicine was established in the year 2000. It was designed to facilitate research-oriented careers in academic medicine for graduates of MSTP and other joint degree programs by providing mentoring and a structured path that would allow the graduates to complete necessary clinical and research training.
Washington University’s Program
The PSTP is centered in the John Milliken Department of Medicine and focuses on the training of internists and dermatologists.5 For individuals interested in internal medicine, the PSTP consists of two years of internal medicine residency training followed by clinical subspecialty fellowship training of one to two years (depending on the subspecialty) and three years of postdoctoral research training (List 1). The dermatology program includes one year of medicine internship, two years of clinical dermatology, and then three years of postdoctorate training.
The participants in the PSTP are encouraged to “fast track” by limiting their training in general medicine to two years. According to the American Board of Internal Medicine (ABIM), individuals who fast track are required to do two or three years of advanced research training after their clinical subspecialty training to become board eligible.6 In other words, the one year of clinical training in internal medicine that is avoided by fast tracking is replaced by two years of research training during the subspecialty fellowship period. This is not a burden on the members of the PSTP because they will be spending at least three years in the laboratory. We feel that the fast track mechanism provides a means to shorten the overall length of the training period without significantly compromising the clinical experience.
The one situation for which fast tracking is not recommended is when the PSTP member is undecided about the subspecialty area he or she wants to pursue. Thus, to fast track, the candidate must declare an area of subspecialty interest sometime in the fall of the internship year. This is not an issue for most members of our program, but it has arisen on a few occasions. When this occurs, the individual is advised to take the time needed to select the most appropriate subspecialty area, even if it adds an additional year of clinical training. There have been two individuals who opted to take three years of training in internal medicine even though they had already decided on a subspecialty area. The PSTP is flexible and allows each member to select the path that best fits his or her needs.
The members of the PSTP are guaranteed a slot in any subspecialty area they select. They must complete the clinical requirements for subspecialty training in toto, as prescribed by the ABIM. For many of the subspecialty areas, this involves 12 months of full-time clinical training followed by 12 to 24 months of weekly half-day outpatient clinic participation. The gastroenterology program requires 18 months of full-time clinical training, whereas cardiology requires 24 months of such training.
After completion of the full-time clinical training, the PSTP participants begin their postdoctoral research training. An important aspect of the PSTP is that participants are not required to perform research in their clinical division or the department of medicine. Rather, they can choose to work in any laboratory at Washington University. Several members of the PSTP have selected laboratories outside the department of medicine, including labs located at the undergraduate campus. This option ensures that trainees will have a wide diversity of research opportunities and allows them to enter new areas that they can subsequently bring back to their subspecialty. The salary of the PSTP trainee is supplemented during the three years of postdoctorate training (currently $15,000 per year) to remove incentives for participating in “moonlighting” activities. The base salary of trainees in the research laboratory is covered either by the clinical division or by the research mentor by use of grants or unrestricted funds from endowment income.
Throughout their tenure as members of the PSTP, individuals meet with the codirectors of the program twice a year. At these sessions, participants are given advice about their clinical and research training options, about potential granting programs, about their specific grant applications, about manuscript submissions, and about faculty placement opportunities. During the academic year, the PSTP holds a monthly session in which participants who are working in laboratories discuss their research. The PSTP also sponsors a yearly half-day symposium at which all the members who are carrying out research give short presentations. This event and the accompanying dinner are attended by an invited physician scientist who also delivers medical grand rounds. The PSTP does not currently include formalized programs in the regulatory aspects of laboratory research or the ethical conduct of research, because these are included in every subspecialty fellowship program at Washington University. Furthermore, the PSTP does not have a formal research program for trainees during their clinical years. However, several PSTP trainees have participated in the C-STAR or Mentors in Medicine programs during their years as internal medicine residents. These programs offer brief clinical and translational research experiences with academic faculty in the department of medicine.7,8
Recruiting for the Program
Given the highly specialized nature of the PSTP, the pool of appropriate candidates is limited, with graduates of MD–PhD programs accounting for the majority of qualified applicants. Graduates of MD–MS and MD–MPH programs are also considered for admission. Another source of applicants is graduates of foreign medical schools who then obtain PhDs at institutions in the United States. Finally, graduates of MD-only programs may be suitable candidates for the PSTP if they have substantial research experience with documented publications. All together, the program currently receives about 80 to 90 applications per year and invites 30–40 of the applicants for interviews.
The PSTP is linked to the internal medicine residency program based at Barnes-Jewish Hospital. When an individual with a combined MD–PhD degree or an MD with a significant research exposure and publication record applies to the internal medicine residency program, they are invited to apply to the PSTP. Some applicants learn of the program through its Web site, which is linked to the internal medicine site. Others hear of the program via word-of-mouth discussions.
In evaluating an applicant, future research potential is the prime consideration, as measured by the publication record, letters of recommendation by those familiar with the applicant’s research, the personal essay, and the interview. Clinical competence and general performance in medical school are also important. At present, the rankings of the PSTP applicants are included within the categorical internal medicine Match process, and the program is not specified by a separate Match number. Every year, approximately 12 to 15 PSTP applicants are listed at a high position on the categorical internal medicine match list for Barnes-Jewish Hospital at Washington University Medical Center.
Outcomes to Date
As of August 2008, 48 individuals have been accepted into the program since its beginning in 2000. All have had combined MD–PhD degrees. Nineteen of the entering PSTP trainees were graduates of the Washington University MSTP, and 29 were graduates of other programs. The target class size is five to seven trainees per year, and during the past seven years the PSTP has averaged 5.0 (SD = 2.3) entering trainees per year. Sixteen have completed (or are about to complete) the program, 25 are at various stages in their training, 6 have transferred to other institutions, and 1 has left the program after internal medicine residency to practice as an academic internist.
The PSTP members have selected subspecialty training in cardiology (11), hematology–oncology (9), rheumatology (5), gastroenterology (3), infectious diseases (3), dermatology (3), and pulmonary (2). Five PSTP trainees have recently joined the program and have not yet selected subspecialty fellowship programs.
Among the 16 members who have completed the program, 13 have remained in academic medicine with appointments as assistant professors of medicine, 1 accepted a position in the pharmaceutical industry, and 2 entered private practice. Of the 13 with academic appointments, 9 joined the faculty at Washington University (2 in hematology–oncology, 2 in infectious diseases, 2 in cardiology, 1 in gastroenterology, 1 in rheumatology, and 1 in dermatology). The others went to the University of Iowa Carver College of Medicine (in pulmonary medicine), the University of Kansas School of Medicine (in rheumatology), the University of Florida College of Medicine (in oncology), and Saint Louis University School of Medicine (in cardiology).
Graduates of the PSTP who remained in academic medicine have obtained a variety of career development awards, including four NIH Clinician-Scientist Awards (K08), two American Heart Association Fellow-to-Faculty Transition Awards, a Burroughs Wellcome Fund Career Award for Medical Scientists, a Pew Scholar Award in the Biomedical Sciences, and a Howard Hughes Medical Institute Early Career Physician-Scientist Award. No graduate of the PSTP has been awarded an NIH RO1 grant, but none of the career development awards have expired yet.
The few who transferred to other institutions before completing the PSTP did so for social reasons, including spousal/partner relocation and child and parental care issues.
Thoughts About the Program So Far
Although the PSTP at Washington University is still in its early days of operation relative to the long training period required to complete the program, the initial results are most promising. Of the first 16 people to complete the program, 13 (81%) have assumed research-based academic positions, both at Washington University and elsewhere. One other graduate accepted a position with a pharmaceutical company, and only two individuals have entered private practice. Based on our tracking of the current members of the PSTP, it seems that at least 80% and likely more will remain in academic medicine as independent investigators. The program has already had a major impact on our department of medicine, with nine of the graduates joining the faculty. A similar outcome has been observed for the PSTP at the University of California, San Diego School of Medicine (UCSD), where five individuals have completed the program since 2003, and all five have assumed research-based academic positions, including three at UCSD (Mark H. Ginsberg, MD, professor of medicine at UCSD, personal communication).
A major issue in the training of physician scientists is the length of the clinical training required for clinical competence and board eligibility. Counting the last two years of medical school, which are clinical, MD–PhD graduates spend a minimum of five years in clinical training between the time they complete their PhDs and start their postdoctorate research training. For some subspecialties, including cardiology, pulmonary, and gastroenterology, the requirements for full-time clinical training are even longer to achieve subspecialty board eligibility. This is discouraging to many MD–PhD graduates, who are anxious to resume research training as soon as possible after graduation from medical school.9 As training programs lengthen, special consideration needs to be given to this subset of physician scientists who play such a key role in the advancement of medical research. For example, physician scientists could be permitted to complete their subspecialty clinical training in a truncated fashion and receive board eligibility for some, but not all, aspects of general subspecialty practice. In cardiology, a physician scientist could be trained to meet the ABIM requirements for the performance and/or interpretation of consultative cardiology, electrocardiography, stress testing, and echocardiography without being trained to meet the requirements for the performance of cardiac catheterization. Because cardiology fellows typically spend six or more months during their two-year clinical fellowship in the cardiac catheterization laboratory, this accommodation could substantially reduce the length of time for cardiology fellowship training. Similar accommodations could be made in other procedure-oriented subspecialties to shorten the length of clinical training.
Another issue concerns the training of MDs who lack PhD degrees for laboratory-based medical research. Our program has focused thus far on individuals with substantial research experience. However, it is well documented that medical school graduates with limited research backgrounds who decide to pursue biomedical research careers can be extremely successful. These are the so-called “late bloomers,” and this category includes many notable scientists, including more than 60% of the Nobel Prize winners in physiology and medicine.10 At this point, we do not accept such individuals into the program because of our limited resources. However, we feel that a goal for the future should be a mechanism for including late bloomers in the program.
Programs with features similar to those of the PSTP at Washington University now exist at several institutions around the country. Most of these programs are focused on the training of academic internists, although some, such as our program and the program at UCSD, also include the training of academic dermatologists.11 Furthermore, the University of Iowa PSTP also includes the training of pediatricians and pathologists for academic positions.12 Hopefully, these programs will expand in the future to include the training of physician scientists entering other areas such as neurology, psychiatry, obstetrics–gynecology, radiology, urology, and surgery. Indeed, Northwestern University has already established a PSTP that included training in all of these specialties.13 The spread of these types of training programs will hopefully increase the number of successful physician scientists in the United States.
The authors thank Mark H. Ginsberg, MD, for providing information about the outcomes of graduates of the University of California, San Diego School of Medicine Physician Scientist Training Program (PSTP). The authors also thank the Edward Mallinkrodt, Jr. Foundation for providing financial support for the PSTP at Washington University School of Medicine.