Strategies for Training Generalists and Subspecialists

The Dilemma

The graduate medical education establishment faces the dilemma of how to change graduate medical education programs to meet the demands of society and at the same time maintain and nurture programs to train the teachers and scientists of tomorrow. This article describes one relatively new (and probably naive) department chair's attempt to develop a solution for one institution. The solution that I proposed to my colleagues at the University of California, Los Angeles (UCLA), is an experiment based on the hypothesis that a positive environment for attracting trainees to careers in general internal medicine can be coupled with a program that attracts a smaller number of trainees to research careers without attracting trainees into careers in the practice of the subspecialties of internal medicine.

The setting for this experiment is a university medical center in a metropolitan area that already is living with a chaotic form of managed competition. Currently, patients with indemnity insurance constitute only 10% to 15% of those hospitalized at Cedars-Sinai and UCLA Medical Centers. The Legislature of the State of California recently passed legislation mandating that half of the University's residency positions be filled by primary care trainees. The governor vetoed this bill on the grounds that the medical schools will find a way to implement this goal without the legislation. However, the bill was reintroduced this year in the State Assembly, where it passed by an overwhelming majority. The legislation did not get out of committee in the Senate, but the author of the bill seems determined that such legislation become law, and he intends to vigorously pursue this issue. Adding to the problem of the immediate need for more generalists and fewer subspecialists is the perception that many subspecialists being trained for academic careers are poorly prepared to compete for research funding with PhD investigators.

The Experiment

Creating a Positive Environment for Primary Care

When I started my duties as Senior Chair on 1 July 1992, approximately 10% of the 325 full-time faculty at the UCLA Westwood campus were in the divisions of general internal medicine and geriatrics. Of these faculty members, only 4 were full-time clinician-teachers; the others were mostly health services researchers. The paucity of full-time clinician-teachers in general internal medicine at UCLA was and still is characteristic of most academic medical centers. To change the balance to one more favorable to attracting trainees to general internal medicine and geriatrics, the department of medicine at UCLA Medical Center, with the blessings and help of the director of the medical center, the chief executive officer of the practice plan (who was convinced that primary care contracts were essential to the economic success of the medical enterprise), and the Dean, have committed to recruit approximately 25 clinician-teacher-internists (including geriatricians) as new full-time faculty members over the next few years. This will enable us to provide capitated care for a sizable population on the Westwood campus at UCLA Medical Center. We believe that these patients and the clinician-teachers caring for them will constitute a prominent and positive image of general internal medicine that will help to entice students to choose general internal medicine and geriatrics as a career.

Economic Incentives

Beginning with the 1 July 1993 academic year, the department of medicine negotiated starting salaries for clinician-teachers in general internal medicine that were more than 10% higher than the starting salaries for board-eligible, catheter-trained cardiologists and other subspecialists starting an academic career at the medical center. We believe that this is an important message to send to our students and trainees.

Noneconomic Incentives

We believe that LaCombe [12] and Barondess [3] were right in stressing the importance of internists participating in the care of their very sick patients. We believe that training a cadre of general internists to care for their patients throughout their illness with less need to consult subspecialists will make the choice of general internal medicine more attractive. We also believe that general internists can be taught common outpatient procedures such as skin biopsies, endometrial biopsies, colposcopy, and flexible sigmoidoscopies, and inpatient procedures such as the placement of Swan-Ganz catheters. Internists can then incorporate the procedures into their practices in both the outpatient and inpatient settings without having to complete subspecialty training. Therefore, we are funding a fourth year of general internal medicine residency training that will be available to a selected few of our trainees. During this year they will choose two subspecialties and, in addition to expanding their knowledge base in these areas, they may choose to learn common procedures for those subspecialties. Those entering this additional year of training will be guaranteed a faculty position after they have successfully completed the training; if qualified, they will be certified to do these procedures in our medical center. The subspecialty divisions responsible for these procedures will be required to incorporate these new “comprehensive internists” into their activities so that the internists will have sufficient ongoing experience with the procedures to be recertified periodically.

It should be strongly emphasized that this additional year of training will not be limited to learning procedures (which is an option for trainees rather than a requirement) but will include additional training to give the trainee two areas of emphasis within general internal medicine. Some trainees may elect subspecialties that do not require procedures, such as infectious diseases or endocrinology. The goal is not to train mini-subspecialists but to create a cadre of comprehensive internists that can entice students into choosing general internal medicine. Moreover, if we succeed in reducing the number of subspecialists, these comprehensive internists (who will spend most of their time in the ambulatory setting) will reduce the need for subspecialty consultation.

Reducing the Number of Subspecialists Entering Practice after Training

We recognize that our efforts to reduce the number of subspecialists entering private practice after training at UCLA will not significantly reduce the number of practicing subspecialists. However, we believe that it is morally and ethically correct for us to begin with our own program and not to wait for solutions to be provided by others. Consequently, after the faculty, the department chairs at the affiliated hospitals, the division chiefs, the Dean, the medical center director, the chief executive officer of the practice plan, and the housestaff were consulted, it was announced in early 1993 that effective 1 July 1994, the department of medicine at UCLA Medical Center will not provide fellowship support to divisions that cannot show that at least 50% of their graduates over the past 5 to 10 years have entered academic medicine, or have made a major change in their programs so that this goal is likely to be achieved. Furthermore, having established this principle, it was also announced that, starting in July 1996, the department of medicine would not allow divisions to accept fellows for clinical training regardless of their sources of support unless they could show that they have achieved or will be likely to achieve this goal by 1996. (It should be noted that when these decisions were made, no data were available on which divisions would or would not be able to meet these criteria).

Better Trained Academic and Research Subspecialists

To better prepare our academic subspecialists for research careers, we have established a novel training track analogous to the Medical Scientist Training Program. We will first encourage our subspecialty divisions to use the Clinical Investigator Pathway of the American Board of Internal Medicine for most of their fellows, with a goal of having most trainees in this pathway by 1996. The time devoted to research during the fellowship will be lengthened to an average of 3 to 4 years in addition to the required years of clinical training for subspecialists. This will effectively and productively increase the time for training subspecialists and will undoubtedly reduce the number of trainees who choose this pathway. In any event, we believe that additional research training will make the graduates of our subspecialty programs more competitive for obtaining research funding.

The Specialty Training and Academic Research Program

To accomplish these goals, we have instituted a new program conceived and directed by Linda Demer, MD, PhD, who has appointments in both the departments of medicine and physiology. Dr. Demer has organized a program with the departments of biological chemistry, microbiology and immunology, molecular genetics, molecular pharmacology, experimental pathology, physiology, faculty members in health services research, and the Rand Corporation that will allow a trainee to take graduate school courses and apply for candidacy for an advanced degree at the masters or doctorate level with one of these departments. Called the Specialty Training and Academic Research Program, it has been given the acronym STAR Program. Individuals can enter the program as medical students, house officers, or fellows. Recruitment for the program will begin in medical school, where merit-based scholarships will be offered by participating subspecialty programs. At the end of the second year of medical school, students who are willing to commit to an academic career in health services research, basic research, or clinical investigation in a subspecialty area will be allowed to compete for a merit-based scholarship. If selected, their third- and fourth-year tuition will be paid and they will be guaranteed (on successful completion of medical school) that they will be ranked sufficiently high in the match to ensure an internship and residency in the department of medicine at the UCLA Center for the Health Sciences. Moreover, assuming satisfactory performance as house officers, they will be nominated for the Clinical Investigator Pathway of the American Board of Internal Medicine. Their research experience through the Specialty Training and Academic Research Program will, if they desire, enable them to be candidates for an advanced degree and to be eligible for board certification in the sponsoring subspecialty upon completion of their program.

It should be noted that students will not be required to pay back the scholarship, that no legal contract to enter a particular subspecialty will be involved, and that students will not even be asked to make a binding commitment to the UCLA residency program. Rather, the program is envisioned to encourage participation by the rewards of special mentoring and by financial incentives such as the scholarships and additional stipends to help offset the cost of student loans during the long years of training required in the program. It should be emphasized that this will be a highly individualized program and that whereas students interested in clinical investigation will participate in a first year of basic graduate school courses, subsequent years will involve study design, natural history of diseases, biostatistics, and other skills necessary for clinical investigation. We believe that there is an increasing need in the biotechnology and pharmaceutical industries for researchers with this kind of training, and we feel that this will be an attractive alternative to academic medicine or practice. It should also be noted that the program is designed to prepare the trainees for careers as independent investigators whether or not they choose to earn an advanced degree. The philosophy and policy of the department is that it will no longer knowingly train subspecialists for practice but will work to ensure that the academic and research needs of society are met and that the best of subspecialty medicine is preserved.

Evaluating the Experiment

I have asked the health services researchers in the division of general internal medicine at UCLA to help evaluate this experiment. We will need to assess the quality and quantity of our recruiting efforts for clinician-teacher internists, including the new comprehensive internists. We will need to evaluate the skills of the comprehensive internists, their success at applying these skills, their effects on their colleagues in the division of general internal medicine and on the subspecialties, their acceptance by patients, and their effect on the cost of providing care. We will also need to evaluate our success in attracting more students into general and comprehensive medicine and in reducing the number of trainees and graduates of the subspecialty programs who enter practice, as well as assessing the success of graduates of subspecialty programs both in finding faculty positions and in obtaining grant support. Finally, we will also have to evaluate the success of all these measures in controlling the cost of health care at our institution while continuing to provide high-quality service to patients.

Initial Reactions to the Experiment

The essence of this proposal was enunciated in an address that I gave to the Association of Professors of Medicine in March 1993. The initial reaction of the housestaff was largely and predictably negative. They were concerned with the bleak picture that I painted of the future for practicing subspecialists. Interestingly, they were concerned with the proposal for training comprehensive internists, specifically teaching them procedures. The faculty in the department of general internal medicine had similar concerns about the program for the comprehensive internist, and the faculty of the subspecialty departments were shocked at the possibility that their fellowship programs could and would be dismantled if they failed to train academicians and researchers. They were also concerned that the mere discussion of such changes would lead to a decrease in applicants for their fellowship positions (this, in fact, does not appear to be the case). Regional and national subspecialty procedure-based societies also expressed concern that the quality of medical practice would be lowered if general internists were trained to do procedures.

However, as the divisions began to collect the outcome data on their trainees regarding the choice of an academic or research career or both, it became clear that all but one of the divisions had met the minimum requirement of 50% of trainees entering an academic or research career over the past 5 to 10 years. The exception, a division whose rate was 41%, has subsequently reorganized its fellowship training program to guarantee that it will be in compliance. The other divisions have similarly realized that the rules have changed, and have altered their selection procedures for fellowship to further enhance their chances of training academicians and researchers. The divisions have begun to consider how to do necessary services without fellows, and have committed to the concept that fellowship is for the training of academicians and researchers, not for service. Recognizing the increasing cost of better and longer training of fewer subspecialists for an academic or research career, many of the divisions have made plans to reduce the number of slots in their fellowship programs.

As a result of a series of meetings at which these issues were discussed in great detail, most of the faculty came to realize that we must accept our responsibility to train generalists and scientists and stop training practicing subspecialists. As the discussions took place, as the faculty became convinced of the correctness of our policies, as the news from Sacramento and Washington, D.C., continued to bear out our predictions (as did the reports back to the housestaff from their colleagues seeking positions after completing subspecialty training), and as they heard of the higher salaries for generalists, the housestaff has dramatically shifted its attitude to one of support. The results of the match in general internal medicine announced in June 1993 were the best our institution has had in many years. The Specialty Training and Academic Research Program has begun to bring the basic sciences, health services research community, and the clinical sciences together in a way that heretofore was not possible. The subspecialty divisions have become highly enthusiastic for the program, and prior to advertising the program we had a surprising number of students, house officers, and fellows who asked to apply before we had been able to finalize our application procedures.

At the October 1993 meeting of division chiefs, it was agreed that within the next 5 years it was very likely that 80% of the trainees in the department of medicine at the UCLA Center for the Health Sciences would be men and women training for practice in general internal medicine; the remaining 20% would be training for careers in research or academic medicine or both. Moreover, the clear consensus was that the only subspecialty training that we should and likely would provide would be through the Specialty Training and Academic Research Program.

Potential Pitfalls and Outcomes for the Experiment

This experiment is predicated on being able to fund a vigorous program in general internal medicine and the Clinical Investigator Pathway for academic subspecialists. These are by no means trivial issues, but I believe they are soluble. To be successful, the experiment will require the cooperation of the general internal medicine faculty, as well as the subspecialty faculty. Their cooperation is likely, given the national and local climates mandating change and the initial reaction to our policies. This reaction is the outcome of the many in-depth discussions of these issues in the past 6 months. The certifying bodies in internal medicine must also help by developing the flexibility to achieve the goals outlined here. In addition, the experiment will require flexibility in various certifications.

The sound of the guillotine [2] is growing louder. We can no longer ignore it or hope that it will simply go away. I believe that if successful, the UCLA experiment will provide a useful model for training internists and subspecialists at academic medical centers located in large urban areas. If the experiment is successful, I will be happy to report to the readership of Annals its progress from time to time should the editors wish me to do so. If the experiment fails, they can invite my successor to report on how and why it did so.