Listen, now, and attend to what I say… When you have done all this, or seen it done, it will be time to ponder… Look to yourself; remember what I told you… he took in thought the course that Athena gave him. - —HOMER, The Odyssey1
In ancient Greek mythology the goddess Athena advanced the concept of the tutor or mentor. Athena disguised herself as a man called Mentor, to school Odysseus. He learned his lessons well.
In the 21st century, we no longer need to disguise our teachers, and our students are doing better than ever. However, the forecast for another generation of talented physician-scientists is almost as challenging and complicated as was Odysseus's journey. There is growing concern that students in the clinical-investigator pipeline and the number of physician-scientists currently being trained in U.S. academic health centers will not be sufficient to continue the rate of current progress in biomedical research. Thus, it is essential that clinicians and scientists establish ways to foster the training of patient-oriented investigators who will continue the current rate of progress and accomplishment in medicine. Furthermore, educators and administrators must develop an infrastructure that will support academic health centers in this research imperative. We wrote this article to outline our views on how these two goals can best be accomplished.
THE LOOMING CRISIS
In 1991, the president of the National Academy of Sciences and the chair of the National Research Council appointed a committee to address “issues in the education and training pathways for individuals pursuing careers in clinical investigation.”2 This panel concluded that training and support for clinical research professionals was “fragmented… undervalued… and potentially underfunded”; there were few programs that offered physicians good preparation for clinical research; and market penetration of managed-care organizations would seriously compromise the translation of fundamental research into health care applications.2
The committee made a series of recommendations to address these issues: multi-year support for clinical investigators through the federal appropriations process, academic institutional recognition of the importance of clinical investigators, changes in graduate medical education accreditation to facilitate clinical research, and the establishment of new coalitions of special-interest groups to identify “investments” in clinical research.2
Although the National Institutes of Health (NIH) and some professional societies have established clinical academic awards, many issues cited in the committee's report remain unresolved. Indeed, JAMA dedicated its July 16, 1997, issue to articles and commentaries on the fate of clinical research.
Medical school faculty traditionally have supported education and clinical research by transferring a portion of patient-care income to the school in the form of an academic pledge, or “dean's tax.” As clinical revenues continue to contract, so does the distribution of resources to both education and research. Thus, the academic health center's ability to fund necessary pilot studies is hindered, making it more difficult for clinical research faculty to compete successfully for national awards. Weissman et al.3 suggested that activities supported by surplus clinical revenues, endowments, institutional reserves, and faculty-funded resources (e.g., devoting discretionary funds and personal time) are becoming more limited. Consequently, internal sources for translational research funding at academic health centers are becoming more difficult to find, and thus the most vulnerable component of the research enterprise, patient-oriented clinical research, may suffer.
In our 1997 article, “Preventing the Extinction of the Clinical Research Ecosystem,”4 we issued a national call to action to address the fragility of patient-oriented clinical research. As a response to this “call,” groups directly concerned (patients, advocacy groups, purchasers of health care, managed care professionals, and other members of the public and private sector) were “summoned” to the Wake Forest University Graylyn Conference Center by the American Medical Association, the Association of American Medical Colleges, and Wake Forest University. The clinical research summit was convened to address a “crisis” that threatens to slow the pace of medical research. During the Graylyn Clinical Research Summit process, over 187 health and business professionals and patient advocates discussed and identified obstacles to the growth of clinical research, including infrastructural and policy barriers to training and career development for clinical investigators.
There is consensus that declining funding for clinical research and training is doing harm to most academic health centers. According to a consortium of pediatrics professional groups,5 in 1999, only 25% of medical school departments of pediatrics held more than ten NIH awards, and 28% of pediatrics departments had no NIH grant. This situation could slow dramatically the present pace of advances in the prevention and treatment of children's diseases. Furthermore, based on exchanges with other academic health center deans, we believe that this “retraction” trend is prevalent in most clinical departments. Thus, the needs of current and future clinical investigators must be addressed expeditiously.
Future physician-researchers face an environment of medical and technical progress amidst continued economic change. Medical progress, however, will depend not only on economic factors but on the quality and number of physician-researchers available to sustain the current pace of biomedical advancement. Academic health centers must assure a continuum of educational opportunities and recruit and retain patient-oriented physician-researchers who “translate” basic research findings to diagnostic, treatment, and preventive modalities.
In the rest of this article, we propose a two-pronged approach to this looming crisis. First, we describe innovative educational programs implemented at one academic health center from the K-12 level through the medical school curriculum. Second, we discuss programs that have been developed to facilitate the recruitment, training, and retention of physician-scientists in the early parts of their professional careers. Four models of training “translational” investigators are presented, along with case studies of how these models have been implemented to create real-life productive and professionally satisfying collaborations within one academic health center.
INNOVATIVE EDUCATIONAL PROGRAMS
Encouraging the Pipeline Approach
At the Wake Forest University School of Medicine and its associated teaching hospitals and clinics, the educational pipeline begins with high-quality math and science instruction in elementary and secondary schools (Figure 1). We believe that the “old” national model for preparing students for medical research careers is guided by chance or serendipity. An enterprising K-12 student or his or her caregiver may find a career advisor or participate in a student counseling service that recognizes aptitude, ability, and, importantly, interest in medical research. The process is one of random chance: chance of connecting with a mentor, chance of finding a role model, or chance of establishing a formal pathway toward a career goal. To encourage high-quality preparation, and establish a more directed approach to career development, our institution initiated a new model, which starts with a consortium of a local school system and a historically black university. The program intent is to develop a generation of diverse, well-trained science students, science teachers, and physician-scientists through early and continued intervention. Through this new-model consortium, two innovative programs—Supporting Teachers and Reforming Systems (STARS) and the Center of Excellence for Research, Teaching and Learning (CERTL)—were created.
STARS, begun in 1997, provides continuous professional development for teachers, using an undergraduate college curriculum designed to cultivate a high-caliber teaching corps, and a support system for increasing the retention of high-quality teachers.6
The mission of CERTL, begun in 1996, is to enhance the quality of teacher preparation through problem-based learning.6 The program supports activities to enhance student performance, focusing on closing the performance gap in math and science for underrepresented minorities.
Problem-based learning emphasizes students' ability to master inquiry and problem-solving skills, which then become transferrable to all learning throughout life. Through specially designed case histories, K-12 students can participate in engaging classroom experiences that motivate them to learn more about topics that are relevant to everyday life experiences and the excitement of medical research. The small groups integral to the model excite the students and broaden their learning and understanding much more than is the case with a lecture-based curriculum.6 We see these activities as benefiting the student's understanding of the fundamental concepts of research. By the end of 1999, over 300 teachers from eight North Carolina counties and from five other states had participated in the CERTL program, which has reached over 1,000 students in the western triad region of North Carolina.
Incorporating Research Experiences into the medical School Curriculum
Studies support the claim that if physicians are exposed to laboratory or clinical research experiences during their student days, they are more likely to carry out research activities in their postgraduate careers.7 We strongly recommend that deans encourage students with college research experiences to continue to learn about clinical research during medical school.
In addition, medical school curricula must provide students with opportunities for meaningful research experiences, especially patient-oriented research. Our institution's new curriculum was designed to do just that. The faculty and administration encourage clinical research and discuss opportunities with matriculants at their orientation to medical school. Throughout the preclinical years the integrative curriculum encourages student involvement in research. A highly competitive NIH-supported program for short-term research training permits over 20 medical students per class to gain experience in both laboratory and clinical research.
We now require every fourth-year medical student to write a scholarly paper prior to graduation. We believe that critical thinking is a prerequisite for the practice of medicine and that the exercise of preparing such a paper not only helps prepare students for the rigors of lifelong learning, but encourages many to consider careers in clinical research.
We remain, however, concerned that U.S. medical student education is not comprehensive enough when addressing issues related to the continuum of research. In a recent editorial8 we suggested that the next edition of Learning Objectives for Medical Student Education: Guidelines for Medical Schools (the first report of the Medical School Objectives Project of the Association of American Medical Colleges) state that medical students should have “a knowledge of the process and values of patient-oriented research, which requires hands-on participation with a human subject.”2 This important competency, though today largely unheeded, was first described in 1991.2
Educational Opportunities for Clinical Faculty
The training of clinical faculty to be better prepared for academic careers is a responsibility of the medical school administration. Although “institutional mentoring” is not classic one-on-one teaching, there does need to be an environment in which faculty can enhance their skills for clinical research and become successful clinical investigators. Such opportunities not only enhance the clinical research enterprise, but also aid in the recruitment and retention of clinical scientists.
After achieving the MD degree, the physician-investigator continues life-long learning in patient-oriented research. At Wake Forest, we foster this process by offering the master's degree in epidemiology and health services research to faculty and medical students. Our goal is to prepare medical students and clinical faculty to enter clinical patient-oriented research with a basic understanding of biostatistics, epidemiology, outcomes research, and other public health issues. Enrollees take 42 credit hours, allowing for continuation of other professional activities. Medical students may enroll at the time of matriculation and graduate in five years with a combined MD-MS degree.
We recently established the NIH-funded Program in Clinical Research to coordinate and integrate physician-investigator training. This core program features a unified and synergistic structure that not only will confer a master's degree in molecular medicine, but also has another track offering an MS degree in health services research and epidemiology. (For more about this program, see the sidebar entitled “Program in Clinical Research.”) A key resource of this program is the NIH-supported General Clinical Research Center (GCRC). The GCRC both provides the infrastructure to encourage patient-oriented research and gives experienced physician—investigators and “graduates” of the other career-development awards the opportunity to interact. This program centralizes the collaborative faculty teaching effort and promotes an optimal learning experience for trainees and mentors.
FOUR NEW TRAINING MODELS
Below, we present four academic-institutional models of physician—investigator training and development for clinical research and suggest ways to enhance the academic infrastructure necessary for another generation of translational research accomplishment. These models are illustrated in Figure 2.
The MD—PhD Model (Model I)
Articles in the January 6, 1999, issue of JAMA, examined some of the contemporary issues related to the MD—PhD in fulfilling the mission of translational research. An editorial by Alison Huang9 proports that while there remains “prestige” in the dual-degree track, a singular “vision of its purpose and necessity is lacking.” The MD—PhD dual-degree model (Figure 2) was designed to train investigators who could “better bridge the gap between basic science and clinical research.”10 Thirty-two Medical Scientist Training Programs (MSTPs) funded by the National Institute of General Medical Sciences (NIGMS) support approximately 900 dual-degree students.10 While most MD—PhD degree holders enter academic careers and contribute significantly to the research enterprise, many select laboratory-based careers over patient-oriented research. A recent study by the NIH to assess the success of MSTP graduates in establishing research careers and their types of careers and research activities confirms this trend.10 Since 1975, the proportion of articles by this cohort published in clinical journals has never exceeded 8%10; this, we believe, reflects a less-than-desirable amount of activity in translational research.
Based on the academic career choices and publication records of MSTP graduates, the NIH, and others,11 conclude, as do we, that these graduates are successful, productive researchers. Huang9 suggests that as medicine and science evolve, so must our nation's dual-degree physician scientists. Our institution believes so strongly in this premise that in the past three years it has re-established an MD—PhD program funded by internal resources. But even with a reorientation to clinical research, we estimate such graduates of MD—PhD programs will constitute only a modest 10% of the physician—investigators needed to conduct clinical or patient-oriented research.
The Mentor Model (Model II)
Academic health centers have seen the need to educate faculty to conduct clinical research. For example, the University of Colorado has initiated a one-year certificate mentoring program in which clinical faculty and fellows take 22 credit hours of course work to develop “basic minimum proficiency and skill” in conducting clinical research.12
Our institution's mentor program pairs a junior faculty member with a more senior scientist, either an MD or a PhD, experienced in laboratory, clinical, or epidemiologic research (Figure 2). The senior scientist will have research-related skills, such as grant writing and the training of clinical researchers. Our program is primarily for clinicians and is divided into two tiers, one for faculty with a departmental commitment to at least 75% protected time for research (the “physician—scientist” group), and one for faculty with at least 35% time for research (the “physician—scholar” group).
In our program, department chairs nominate both mentors and fellows. A medical school oversight committee of senior basic science and clinical faculty coordinates the entire process and has responsibilities for facilitating the mentor—fellow pairing and monitoring their success. Interdepartmental mentor—fellow pairing is strongly encouraged. We are now providing some resources and considering reimbursed protected time for the mentors.
Physicians hired at the end of their training are often not prepared adequately to succeed “on their own.” The scientist-and-scholar programs team such individuals with experienced researchers in a relationship similar to that of a formal basic science postdoctoral fellowship. The program provides the fellow ready access to the mentor and his or her laboratory or facilities. The “laboratory” can include any site for conducting translational research, such as the computing facility or clinical population-based research modules. For such collaborations to be successful, the clinician must understand the fundamentals of his or her research field and be able to participate at the conceptual level. The period of mentoring provides the experience, knowledge, and confidence to facilitate such roles. The mentor receives support to attend courses relevant to mentoring skills. After three years, our experience with 11 fellows has been gratifying. The Department of Pediatrics has sponsored six fellowships in various disciplines, and the first fellows have already obtained NIH funding.
Given today's constrained resources, the commitment for protected research time is significant and needs academic and federal intervention. We propose a NIH-supported translational research protected-time pilot program for physician—investigators who have engaged in or completed formal institutional mentorship programs. Such investigators would be eligible to compete for NIH support for research protected-time salary. The proposed program would require the institution to guarantee the faculty member research-protected time and research resources equivalent to up to 40% of his or her salary. The awards would be merit-based for translational research projects that can be completed in two years. This investment in the next generation of translational investigators may be considered analogous to early-stage investing in promising biotechnology companies. However, we believe the valuation of the physician—investigator is much greater and the anticipated high return—an experienced investigator who is ready to move to the next phase of academic clinical research leadership—is worthy of NIH consideration.
It is too early in our experience, however, to gauge the value of such a program in developing truly independent patient-oriented clinical researchers. We estimate, based on faculty interest and available resources for post-MD career development in clinical departments, that if other academic health centers had similar physician—scientist programs and a NIH protected-time program were initiated, at least another 20% of the necessary clinical investigators would be produced.
Molecular Medicine—PhD Clinical Training Program (Model III)
To meet the challenge of training basic scientists who will function successfully in the clinical environment, several major academic centers, including ours, have initiated graduate (PhD) programs that provide training in the use of cellular, molecular, and integrative approaches to better understand human disease (Figure 2). The goal is to have PhD investigators support, in part, the clinical research enterprise. Often, these programs are located within clinical departments, with faculties of physician—scientists who have major ongoing research programs. These graduate programs differ from others by offering PhD students a more comprehensive knowledge of human biology and the etiology and pathophysiology of disease. This training allows the trainees to develop research programs with clinical implications.
However, these programs have specific limitations. Researchers may be isolated in clinical departments, and they may have limited access to the educational programs of other institutional graduate students, and only loose connections with those scientists with similar concerns or interest. We believe, however, that with new types of collaboration we can counter this isolation. These programs can augment the institution's capacity to recruit and retain outstanding basic scientists to work in clinical departments. They can also facilitate communication between departments and the optimal use of common research facilities currently available in all school departments. We have included this program in our Program for Clinical Research.
Finally, PhD clinical research programs also provide a wider range of research opportunities by enabling clinical (MD) fellows to perform research side-by-side with PhD graduate students. We believe this approach enriches the laboratory experience of both clinical fellows and graduate students. While PhD clinical research programs are an important addition to our school's diverse opportunities for clinical research training, we estimate they will provide only another 15% of the numbers of investigators (and not physician—investigators) needed to meet the pace of clinical accomplishment.
The MD—PhD Pair Approach (Model IV)—The Model for Future Success
An effective model both to promote the development of academically successful clinicians and to continue the pace of research accomplishment is a pairing of fundamental scientists and physician—investigators to carry out research based on a single hypothesis. The MD—PhD pair has, at our institution and elsewhere, proven to be a valuable research resource. The talents of academically trained researchers meshed with the expertise of clinical practitioners form a research team that links fundamental mechanisms of disease research with clinical diagnosis, prevention, and treatment.
Our experience shows that a minimum of 20% protected time for the physician's research and 80% research time for the basic scientist provides one full-time equivalent employee (FTE) devoted to a single-hypothesis research project. This model can, with the support of academic deans and federal funding agencies, provide sufficient support for the continuation of biomedical advancement.
To be effective, the team of physician and fundamental scientist has to be just that—a team. The academic degree should not imply that for translational patient-oriented research, the PhD is a contract employee or a hired hand. The team should be a viable unit, as described below. Recognition of accomplishments by promotion and tenure committees is essential for this model to work. Deans of academic health centers must change the culture and confer upon PhDs who perform their important role in translational research protocols the same value as is conferred upon those who remain in basic science departments investigating fundamental issues. At Wake Forest University School of Medicine we endorse this team approach and have been carefully reviewing the outcomes of promotion committees to assure that the progress and accomplishments of both our translational research team members are recognized.
Here, we summarize three different cases to illustrate the potential success of this model for the future of clinical research.
Case 1: The “serendipitous pairing” of a surgeon with a doctor of pharmacy, thus bringing together professionals from two departments (surgery; medicine)
Origin: This pairing of a general/trauma surgeon who specializes in oncology with a doctor of pharmacy whose research interest is hematology/oncology began when these individuals met at a Cancer Center Grand Rounds.
Collaboration: After an initial meeting a productive collaboration developed, due largely to the basic scientist's interests in patient-oriented cancer research and his understanding of the unpredictability of the clinical research process. This understanding allowed him to accommodate his experiments to the physician's variable “trauma call” schedule, and the unpredictability of human experimental tissue availability. He also accepted the necessity of infrequent investigator-to-investigator consults and discussions. The surgeon was experienced in hypothesis-driven research due to a two-year research fellowship. Both believe their research success is due to their common interests and understanding of the differences in their medical and pharmacy training.
Results: This team has been successful in publishing peer-reviewed clinical research articles, obtaining funding from the NIH, and developing patents for new treatment regimens for patients with particular types of cancer.13
Case 2: The “institutionally activated pairing” of a physician with a scientist holding a PhD, thus bringing together professionals from two departments (cancer biology; public health sciences)
Origin: This collaboration was established when our institution responded to a program announcement from the NIH. The chair of the Department of Cancer Biology initiated the contact.
Collaboration: This team believes they are successful translational researchers because of a mutual respect for each other's expertise and capabilities, which are both complementary and synergistic. The oncologist has 30 years of patient contact and access to many patients. The epidemiologist has grant-writing skills and expertise in hypothesis-driven research. She is, in her words, “methodologically oriented,” with experience in systems approaches to population-based research. Their success is built upon being co-equals in all of their joint ventures.
Results: The team has obtained a number of NIH-investigator—initiated research grants and contracts. They have planned scientific conferences together and written a number of peer-reviewed articles in established medical journals.14
Case 3: The “designed pairing” of a scientist with a PhD in engineering with a physician surgeon, both from the same department (surgery)
Origin: This team was established when a plastic and reconstructive surgeon who wanted to do research realized he no longer had time to be a clinician, a teacher, and a researcher. His solution was to hire a PhD engineer as a laboratory-based fellow who would later become a member of the faculty.
Collaboration: The surgeon's philosophy was that keeping his knowledge base updated in an era of unprecedented progress required him to focus time on clinical issues and rely on outside expertise for research. The postdoctoral research fellow devoted time to basic science advances. Their complementary knowledge bases allowed for continuity between the development of a concept and its clinical application.
Results: The team's success is best illustrated by the patenting, licensing, and introduction of innovative technologies (e.g., a vacuum-assisted wound closure device) that have advanced the science of wound healing.
If clinical research is to be sustained at the present level of accomplishment, we must recognize that health-care-system and economic changes will continue to influence the career development of clinical investigators. The number of new investigators applying for grants between 1985 and 1993 dropped by 54%, and there was a 21% reduction in grant applications from first-time physician—investigators between 1994 and 1998.15 We also believe that the decline in American medical students' interest in clinical research, as shown by Rosenberg,16 may continue, and that the gap of unrealized research opportunities will grow unless action is taken (Figure 3).
We urge all academic health centers to make a commitment to foster all the models we have described above. The underlying premise is that schools of medicine should support the teaching of patient-oriented research methods and practices so that our nation will not lose a generation of clinical investigators. Finally, we believe that the adoption of new models requires that institutions give adequate protected time for physician—scientists to engage in clinical research, even at the modest 20% described for the “pairing” model.
Thus, we implore deans of academic health centers, the NIH, the Agency for Healthcare Research and Quality, and the many organizations considering the future of academic research to develop external and internal mechanisms of financial support to provide institutional encouragement and “protected time” for the present cadre of physician—investigators. We ask the NIH to consider carefully the enactment of a translational research protected-time pilot program.
The scientific rigor that Abraham Flexner espoused, and the resulting growth in excellence of our medical education system, led to the combination of science and medicine upon which our health care system is founded—a system based upon inquiry and discovery, resulting in startling innovations in diagnostic and treatment technologies.
If we are to continue with the great successes of recent decades, we as a nation must consider the process by which we prepare men and women to enter careers in clinical research. Thus, academic health centers must undergo a planned transition from what they were a decade ago to what they plan to be a decade from now, adapting to a new synthesis of initiatives and strategies to enhance translational research.
More than eight centuries ago Rabbi Moshe Ben Maimon (Maimonides) prayed: “Should those who are wiser than I wish to improve and instruct me, let my soul gratefully follow their guidance; for vast is the extent of our art.”17
The scope of science in the 21st century is vast also, well beyond our present knowledge. To be prepared for the effects of new knowledge on human disease, suffering, and death, we must enhance training opportunities for future physician—scientists.
Program in Clinical Research
Wake Forest University School of Medicine, the Graduate School of Wake Forest University, and the Wake Forest University Baptist Medical Center have established a new Program in Clinical Research (PCR) to coordinate and integrate physician-investigator training. The PCR, begun in 1999, is funded by a National Institutes of Health Clinical Research Curriculum Award and internal funds from the university. The goal of the program is to provide education and training in the skills required to conduct high-quality clinical research. The PCR consists of graduate-degree programs that encompass many of the definitions of clinical research: Clinical Epidemiology and Health Services Research, Patient-Oriented Research, and Molecular Medicine. Two programs (Clinical Epidemiology and Health Services Research, and Molecular Medicine) lead to a master's degree. A PhD program in molecular medicine, and an MD-PhD program, are also part of the PCR. A master's degree in human genetics is planned, which will be coordinated with the Wake Forest University School of Medicine's newly established Center for Human Genomics.
The PCR also provides numerous support services for clinical research, partnerships with industry, and stipend programs supported by training grants and scholarships. The ethics of human research, legal and regulatory issues, and technology transfer are emphasized. The PCR has a strong mentorship program.
Research activities with the PCR are multifaceted and involve both basic and clinical departments. Categories of clinical research are the molecular mechanisms of human disease, human genetics, women's health, the biology of aging, therapeutics, epidemiology, health services, and new technology and devices. There is special emphasis for both students and clinical research faculty on education relevant to the conduct of human and animal research and the general field of research ethics. (Formal training in clinical research ethics and policies occurs with a requirement for obtaining from the institution a certificate that reflects a formal written test of competency.) The integrative PCR model will prepare students for careers in clinical research. The PCR's Web site, www.wfubmc.edu/pcr, offers specific information about this new initiative, which is showing great promise in encouraging patient-oriented research.