Encouraging Clinical Research by Physician Scientists

Why is it important that physicians actively conduct clinical research? Well-trained PhD scientists, working with physicians, can often design meaningful experiments to be conducted in human subjects and can collect and analyze data independently or in collaboration with physicians. But only the well-trained physician scientist can thoroughly understand, interpret, and properly care for human subjects during studies that involve an intervention. Even more important, the physician scientist is uniquely positioned to ask the relevant questions that will redefine the therapeutic and preventive opportunities and to identify the human conditions, inherited or acquired, that offer new opportunities to advance health science. The conclusions and recommendations of the National Institutes of Health (NIH) Director's Panel on Clinical Research reported by Nathan et al¹ provide useful additional data and make constructive recommendations to strengthen clinical research and to increase the participation by a wide variety of health professionals and scientists in the clinical research enterprise. The leadership of the NIH has responded with new K23, K24, and K30 awards designed to stimulate the field. These awards and the establishment of a medical student program in clinical research at the NIH are important initiatives, but much more must be done.

Although the panel identified that 36% of the dollars awarded in the NIH extramural program were for "clinical research grants," as they defined them, their own data demonstrate that physicians comprised only 36% of the principal investigators of these grants.¹ Since the majority of experiments that directly involve patients as human subjects are conducted by physicians, these data are quite consistent with the analysis of the 1994 Report of the Institute of Medicine,² which suggested that approximately one third of clinical research grants, and therefore about 10% of NIH extramural funds, supported "hands-on" research involving patients and physicians. The NIH Director's Panel's recognition that 52% of the principal investigators receiving clinical research grants had PhDs emphasized the significance of biostatistics, epidemiology, behavioral medicine, and outcome research as areas for research investment. Indeed, the definition of clinical research should be broad, extending from the study of tissues obtained from humans to the study of human populations. But these data, combined with the 30% decline in the number of new physician applicants for NIH funding between 1994 and 1996, reemphasize the urgency with which the role of physician clinical investigators must be encouraged.

Although the panel's report emphasizes that the changing health care environment, with cost constraints and the policies of managed care organizations, has placed enormous pressure on the clinical research enterprise, it is important to recognize that threats to the clinical investigator were identified 20 years ago, well before managed care was part of the daily health care lexicon.³ While income from clinical practice has been extremely important in the support of clinical research, particularly that conducted by physicians, the concerns about the "endangered species"³ were recognized during a period in which patient care revenues to practice plans and academic health centers were growing. Part of this paradox arose from difficulties inherent in clinical research itself. The complexity of working with human subjects, the long time frame involved in many studies, the requirement for involvement of multiple investigators, difficulties with credit for multiauthor papers, and the myth that any dedicated clinician can become a good clinical investigator contributed to difficulties in supporting and rewarding clinical investigators. The reductionism of research in the bench laboratory allows far more control of the experimental conditions. Scientists devoting 75% or more of their time to research can generate a substantial bibliography in the bench laboratory, as judged by the number of publications. This discrepancy has been further amplified by the difficulty for a study section made up primarily of PhD scientists to evaluate clinical proposals.¹ In clinical training programs, a critically important educational precept—that students learn as much or more from their peers as from their mentors—was largely ignored in clinical research training except in some large institutions. Many successful senior clinicians believe that they are mentoring researchers, when, in fact, young clinicians are often used largely for data collection.

The new K30 award for the first time offers an incentive to create integrated curricula for clinical training. It is essential that the successful awards be located in institutions with strong active clinical research centers, a critical mass of investigators, and a critical mass of students who can interact with each other and learn together. This will require a commitment to break down departmental and divisional lines that isolate trainees. In the long run, new interactions of trainees might produce some exciting new research.

If academic institutions do not promote clinical investigators, recognizing that quality and creativity are essential but that the metrics are quite different from those applied to bench laboratory scientists, the effort will fall short. Successful NIH funding for clinical research will provide one of the metrics for the promotion process, but the same biases that have existed within the NIH study sections must not be perpetuated in the academic health center.

Clinical research is complicated by many obstacles, ranging from institutional review board requirements, Food and Drug Administration approvals, and a host of other procedural challenges. Academic health centers must establish mechanisms to help clinical investigators negotiate these obstacles. Some institutions (Massachusetts General Hospital, for example) have provided important models for accomplishing these goals, but many more should do so. I continue to believe that creating a national pool of clinical investigators' candidates for training and matching them to training sites, similar to the Robert Wood Johnson Clinical Scholars program, would enhance the opportunities to recruit women and minorities for careers in clinical research. It is much more difficult to select trainees of diverse backgrounds for 1 or 2 positions at a single institution rather than from a pool applying for 30 positions at multiple institutions.

The NIH Director's Panel's call for the convening of interested and relevant parties is a useful recommendation. In this discussion, the potential role for foundations should be amplified. Foundations have the capacity to stimulate new projects to interest and support medical students toward careers in clinical research. Particularly needed are additional MD/PhD programs, which allow physicians to obtain their PhDs not only in the bench laboratory but also in clinical research directly involving human subjects, and programs in the social and behavioral sciences, economics, mathematics, epidemiology, and other important aspects of health care and health science.

The resistance to this kind of preparation in the current Medical Scientist Training Program (MSTP) is high. Students interested in careers in clinical investigation or other non–laboratory-based disciplines have been actively discouraged by many program directors and faculty in the existing MSTP programs. The NIH might want to consider some partnership programs with foundations in these areas.

It is quite unlikely that the NIH will restore the Biomedical Research Support Grant program, at least in its previous incarnation. In that life, funds were used flexibly and often awarded without peer review. Many individual scientists (with the exception of those who received some of the funds) often believed the monies were distributed without adequate rigor. The reestablishment of formal institutional funding will require the following: (1) clear articulation of the purposes for which the money is to be used, ie, seed grants, bridging funds, and equipment upgrades, (2) mandatory peer review conducted by funded investigators at the institution, (3) careful accounting of the uses and productivity of the investment, and (4) a new name for the program.

If the new NIH initiatives are to work, there must be a clear demonstration to clinical investigators that adequate funding of meritorious research will be available in the future. It is hoped that the K23, K24, and K30 grants will galvanize interest of young people to careers in clinical investigation. But if at the conclusion of this training meritorious proposals for well-trained individuals are not funded by the NIH at a recognizably significant level, the effort will fall short. This will require effective reorganization of study sections, efforts toward which are under way. The clinical research community also should support efforts by the NIH to obtain legislative approval for a graduated debt repayment for committed investigators to the field.

Clinical research always will depend on many health professionals, including physicians, PhD scientists, nurses, dentists, and social and behavioral scientists, who work together in the conduct of clinical research. The NIH Director's Panel has made a significant contribution and the NIH leadership has initiated important follow-up efforts, but the physician clinical investigator is still on the "endangered species"³ list. Perhaps these are the first important steps in changing that classification.