Teo, Alan R. MD
The demand for training in the conduct of clinical research can be traced back to the increasing complexity and sophistication of medical research design after World War II. One of the earliest large-scale randomized controlled trials (RCTs) was the Salk polio vaccine field trial of 1954. Landmark cardiovascular studies initiated in the late 1970s and early 1980s such as the Hypertension Detection and Follow-up Program Cooperative Group1 and the Physician Health Study2 further helped establish RCTs as a highly effective method for clinical research. List 1 shows a timeline of some of the key events in the evolution and development of clinical research training in the United States. The term “clinical research,” for the purposes of this article, refers to the National Institutes of Health (NIH) definition, which encompasses patient-oriented research, epidemiological and behavioral studies, and outcomes and health services research.3
As the complexity (e.g., RCTs and meta-analyses) and scale (e.g., multicenter trials) of clinical research have grown, so has the technical expertise necessary to execute it. Human participants, trials involving healthy individuals, and involvement of pharmaceutical companies are examples of thorny issues that require conduct of the highest ethical standard. Data manipulation requires a mastery of difficult mathematical concepts in statistics and facility with software to analyze raw data. Writing skills, too, are indispensable in securing funding from grants and institutional review board approval of research protocols. The result is a diverse skill set too difficult to master without formalized curriculum and training.
With the increased focus on RCTs and clinical research, clinical epidemiology, described as the “basic science”4 of clinical medicine, and biostatistics started to move from a peripheral intellectual field to an important presence within medical schools and an indispensable ally in modern clinical medicine. David Sackett founded Canada’s first department of clinical epidemiology and biostatistics in 1967 at McMaster University, and he published a seminal article introducing the concept of clinical epidemiology two years later.5 Public health schools, which also taught the topics, became more linked to medical schools.
Sackett and his colleagues at McMaster University also receive credit for starting the global evidence-based medicine (EBM) movement by making available reliable information on therapies to all clinicians. It was not until 1992, however, that the term “evidence-based medicine” was first widely disseminated in another seminal article that described it as a “new paradigm in medical practice.”6 Medical school and residency programs nowadays include EBM to varying degrees in their curricula.7,8
Tackling a Crisis in the Clinical Research Workforce
By the 1980s, a number of eminent research-oriented physicians publicly warned about an impending shortage of clinical researchers. James Wyngaarden,9 then at Duke University but later head of the NIH, rang the first warning bell in 1979. Using a series of line graphs depicting the decline in NIH grants and fellowships awarded to MD investigators in contrast to the steady rise among PhD investigators, he famously heralded the clinical investigator an “endangered species.” Wyngaarden and other doctors were particularly concerned about the lack of physician–scientists pursuing research that directly involved patients and that could be applied to improving care for their diseases. They suggested that the problem was not so much a lack of interest in clinical research per se but, rather, competition with careers in booming areas of basic science such as molecular biology.10
The 1990s saw more detailed analyses of the problem. Edward Ahrens,11 for one, produced a 200-page book. He was particularly concerned about a gap in the spectrum from basic bench research to applied clinical research, the area of “basic patient-oriented research” that investigates topics such as physiology and the fundamental mechanisms of human disease. Major national organizations like the Institute of Medicine, the NIH, and the Association of American Medical Colleges (AAMC) followed suit with reports focused on clinical research. The Institute of Medicine estimated that just 10% of NIH research at the time was clinical in nature.12 In 1995, Harold Varmus, then head of the NIH, formed a special blue ribbon panel that released the so-called “Nathan Report” two years later. This report was instrumental in fostering more clinical research by offering a series of key, concrete recommendations.13 Among them was the recommendation for awards that would support the development of clinical research training curriculum, realized in 1999 with the introduction of K30 awards. And, the AAMC, though slower in response, issued its own report delineating nine core problems and recommendations, including a specific call for formalized training programs.14 Medical specialties, spurred by NIH recommendations, funded clinical research training grants for junior faculty motivated by an academic career in clinical investigation.15 Concerned about the rapid decline in the number of clinical investigators, a group of physician–scientists committed to seeing medical science advanced by the active participation of real patients and bedside observations by such clinical investigators formed the Association for Patient-Oriented Research in the late 1990s.16
An important backdrop to these increases in funding for clinical research training was the general increase in funding for medical research. From 1998 to 2003, the NIH underwent a boom in funding, thanks to political support, and its annual budget doubled from $14 to $28 billion.17
This allowed the NIH to provide funding mechanisms such as the K awards that have been essential in encouraging the development of clinical investigators. Most K awards provide salary support to clinical investigators so that they can devote themselves to research, a critical concept called “protected time.” The K23, K24, K08, and K12 awards, for instance, are specifically for those pursuing clinical research careers. The K23 award is formally called the Mentored Patient-Oriented Research Career Development Award and aims to develop independent clinical research scientists. It is designed for young faculty members at academic health centers and provides them with three to five years of salary support for protected clinical research time. Importantly, these grants compensate researchers for salary: About $120,000 can be designated per trainee, and, of that, $90,000 can be allocated to salary.18 Though K23 awards began as a minor fraction of K awards, they now compose approximately 20% of the dollar amount of total K grants.19 K24 awards continue the support of K23 awards for midcareer investigators in patient-oriented research. K08 and K12 awards are similar multiyear awards designated for those with a clinical doctorate (e.g., MD or PharmD) interested in developing a clinical scientist research career.
Early Examples of Clinical Research Methods Coursework
Public health schools probably offered the first major educational experience to integrate topic areas essential to conducting clinical research—epidemiology, statistics, and research methods. Decades before departments of epidemiology and biostatistics gained a presence in medical schools, they held a central position in public health schools. Though the training was geared toward those interested in health at a population level, the skill set was still applicable to clinical research.
Thus, many of the early pioneers in developing clinical research training programs were MD/MPH physicians who applied their additional training in public health to patient-oriented research. Nonetheless, there were still limitations to the applicability of master of public health (MPH) training for clinical research. Prime among them was that MPH coursework was just that—coursework. It lacked an applied component, like a practicum, in which trainees would actually design and implement a research protocol.
The Robert Wood Johnson Clinical Scholars Program, begun in 1972 after a pilot program, is an example of a small but well-known program that includes components of clinical research design. It has evolved over the decades, but it has remained true to a goal of training young physicians to conduct research, particularly in the area of health services and outcomes research, to help rectify the imbalance between basic biomedical research and patient- or population-oriented research.20,21 Then as now, it is a two-year integrated educational experience for doctors who have completed their residencies. Participants take coursework in population health, epidemiology, research methods, health care organization, economics, and health policy. They are rewarded with a master’s degree on completion. Training is conducted at major universities’ academic health centers and under a plethora of potential research mentors. More than a thousand scholars have graduated since the program’s inception.
Like public health schools, however, the primary goal of the Robert Wood Johnson Clinical Scholars Program is by no means to merely create clinical researchers. It was designed for physicians interested in leadership roles within medicine who would learn about the big, broad issues in health care. And, like those who earned MPH degrees, Clinical Scholars Program graduates acquired as a fringe benefit knowledge about some aspects in the design and execution of clinical research.
Therefore, although there were some early examples of educational opportunities related to clinical research, programs or classes devoted primarily to clinical research design did not emerge until about the 1980s. Using the search term “clinical research AND (training OR curriculum OR methods),” a PubMed search indicated that the earliest publication to describe how to conduct clinical research dates back to 1980. Two biostatisticians at the University of Washington wrote a practical paper describing a model class in clinical trial design and management.22 The University of California, San Francisco (UCSF) was one of the pioneers, first offering an intense summer course in designing clinical research in the early 1980s. Likewise, texts on clinical research design began to be published in the 1980s.4,23 Duke University was one of the nation’s first to create a master’s degree program in clinical research as early as 1986.15
The Development of Formal Clinical Research Training Programs
One type of K award, the K30, is largely responsible for the development of current clinical research training programs. It provides up to $200,000 per year in funding to institutions specifically for the development of clinical research curricula. The first set of grants expired in 2005, and the second set, many of which have been renewed by the same institutions, will end in 2010. Fifty-one institutions across the United States have received K30 awards.18
This funding, which supports on average 64% of a program operating budget,24 has inspired nearly 60 institutions across the United States to date to develop degree-granting clinical research training programs. Of those surveyed in 2004, 80% of K30 grant recipient institutions offered a certificate, 78% a master of science, 31% an MPH, and 20% a PhD.24 The AAMC has assembled a collection of such training programs with detailed information available online at (http://www.aamc.org/research/clinicalresearch/training/start.htm).25 The institutions that created these programs typically are large universities with major medical schools and university hospitals that share a mission to educate a new cadre of health professionals, from nurses to pharmacists to physicians, interested in careers in clinical research. Through a combination of coursework, practical hands-on experience, and mentorship, students learn all about how to design and conduct clinical research.
Besides degree-granting options, some clinical research training programs offer abridged curricula for clinicians interested in clinical research but too busy to take a year or two to earn a degree. Thus, these nondegree or certification programs still serve an important function: They provide the bare essentials in clinical research training for clinicians who otherwise might not get such training. They also seem to be slightly more popular, as 56% of graduates of K30-funded clinical research training programs earned certificates compared with 50% who earned a master’s degree and 4% who earned PhDs.24
Two illustrative examples of such programs are at those at UCSF and Harvard University. UCSF began offering an intensive two-month workshop in the early 1980s (http://www.epibiostat.ucsf.edu/courses/summerworkshop.html), which more than 100 students complete each summer. Students take a package of three classes: (1) Designing Clinical Research, (2) Responsible Conduct of Clinical Research, and (3) Building an Academic Career. The first class provides a core overview of the chronological process of designing a study, from formulating a research question all the way to writing a short research protocol. The second class focuses on ethical considerations. In the third class, students learn about finding mentors and getting grants and must sketch a two-year career plan. Harvard’s equivalent is called the Summer Program in Clinical Effectiveness (http://www.hsph.harvard.edu/clineff). Participants take two core courses in biostatistics and clinical epidemiology and two electives (such as in quality-of-care research or research with large databases) during the summer. During its history of more than 20 years, about 1,700 clinicians have completed the program.
Waiting until after medical school and residency to start training in clinical research is late. To build a talented workforce of investigators requires fostering an interest in clinical research careers at earlier stages. The AAMC, which oversees curricula nationally in the United States, called for all medical schools to incorporate mandatory education on clinical and translational research in 2006.26,27 Medical, pharmacy, nursing, and dental schools, especially research-oriented ones, may also foster interest by offering scholarships or grants for student research projects or encourage students to use a summer vacation to work with a researcher on an ongoing clinical trial. On a national level, several programs have been developed in recent years that seek to both make students aware of the excitement of clinical research and train some of the brightest young minds (List 2).
Again, the NIH offers a good example of one such program. The NIH Clinical Research Training Program, begun in 1997, is a one-year experience that targets medical and dental students. Because the program is clinical in nature, students must have completed a year of clinical rotations to be eligible. In exchange for delaying graduation by a year, participants receive a living stipend of approximately $24,000. The program, held on the NIH campus near Washington, D.C., doubled in size in 2004 to 30 students annually. It includes a series of classes and seminars, requires participants to conduct research under the guidance of an NIH researcher, and culminates in a conference at which trainees share their results at the end of the year.
Recommendations for Maintaining a Robust Culture of Clinical Research Training
Model clinical research training programs use multiple teaching methodologies and components, as enumerated in List 3. First, they include a traditional core didactic curriculum to establish a theoretical foundation and base of knowledge. Second, they reinforce such lectures with seminars and small discussion groups. This allows clarification of difficult topics. More important, it offers a forum for richly experienced students to share with peers examples of research projects in which they have been involved. Third, a practical experience based on the student’s own research interest is required; participants often write and carry out a research protocol as part of the curriculum. Fourth, each student has an individual mentor. Having a faculty member who carefully and critically evaluates one’s protocol is indispensable. Fifth, these programs should include supplementary workshops on the assortment of tools used in clinical research. For instance, every clinical researcher should be skilled in using bibliography-organizing software such as EndNote, statistical software such as SPSS or SAS, medical literature databases (particularly a thorough knowledge of PubMed), and writing grants. Sixth, the program should result in at least a certificate, if not a degree, as evidence of the rigor of study.
A “build it and they will come” mentality is insufficient, though, to ensure that young clinicians pursue careers in clinical research. To avoid another workforce crisis like that observed in the 1980s, I have assembled a series of eight recommendations along with their theoretical justifications (List 4). It is worth mentioning that the very rigorous evidence basis that is so central to contemporary clinical research is difficult to apply to measures to enhance clinical research training and careers. The limited studies of the impact of programs have tended to be either of qualitative descriptions of what graduates have done20,28 or of programs that failed to find a conclusive benefit.15 One of the most methodologically rigorous studies demonstrated a static number of physicians applying for first major research grant (R01) from the NIH during 40 years, suggesting no increase in new physicians pursuing clinical research careers27; however, the impact of reforms beginning in the late 1990s may not have been observed by 2004 when the study data ended. What these recommendations do represent is a synthesis of empirical strategies that have been used with anecdotal success.
Looking Toward the Future
The NIH again served as the key impetus for change in 2002 when Director Elias Zerhouni began discussions that resulted in the Roadmap Initiative. This huge, ongoing endeavor is designed to accelerate the translation of basic science findings to clinical practice. In particular, the NIH has heavily invested in creating a national consortium of clinical and translational science institutes. So far, 24 academic health centers have received Clinical and Translational Science Awards.29 Many of these grants place specific emphasis on improving educational programming to train the next generation of clinical researchers.
Concerns over the balance between basic and clinical research will doubtless continue in the future. In particular, the subgroup of clinical investigators that is specifically patient oriented as opposed to laboratory based remains an endangered species because of a dearth of funding opportunities. Nonetheless, it is amazing to consider that in the span of little more than two decades, the United States has gone from experts decrying the state of the clinical research enterprise and a few scattered offerings to a nationwide network of institutions, programs, and funding for education in how to conduct clinical research. It is exciting to consider what the next development will be and the impact of training programs in the coming years.
1 Five-year findings of the hypertension detection and follow-up program. I. Reduction in mortality of persons with high blood pressure, including mild hypertension. Hypertension Detection and Follow-up Program Cooperative Group. JAMA. 1979;242:2562–2571.
2 Final report on the aspirin component of the ongoing Physicians’ Health Study. Steering Committee of the Physicians’ Health Study Research Group. N Engl J Med. 1989;321:129–135.
4 Sackett DL, Haynes RB, Tugwell P. Clinical Epidemiology: A Basic Science for Clinical Medicine. Boston, Mass: Little, Brown and Co.; 1985.
5 Sackett DL. Clinical epidemiology. Am J Epidemiol. 1969;89:125–128.
6 Evidence-Based Medicine Working Group. Evidence-based medicine: A new approach to teaching the practice of medicine. JAMA. 1992;268:2420–2425.
7 Green ML. Evidence-based medicine training in internal medicine residency programs a national survey. J Gen Intern Med. 2000;15:129–133.
8 Holloway R, Nesbit K, Bordley D, Noyes K. Teaching and evaluating first and second year medical students’ practice of evidence-based medicine. Med Educ. 2004;38:868–878.
9 Wyngaarden JB. The clinical investigator as an endangered species. N Engl J Med. 1979;301:1254–1259.
10 Gill G. The end of the physician scientist. Am Scholar. 1984;53:353–368.
11 Ahrens E. The Crisis in Clinical Research: Overcoming Institutional Obstacles. New York, NY: Oxford University Press; 1992.
12 Kelley WN, Randolph MA, eds. Careers in Clinical Research: Obstacles and Opportunities. Washington, DC: National Academy Press; 1994.
13 Nathan DG. Clinical research: Perceptions, reality, and proposed solutions. National Institutes of Health Director’s Panel on Clinical Research. JAMA. 1998;280:1427–1431.
14 Clinical Research: A National Call to Action. Washington, DC: Association of American Medical Colleges, American Medical Association, and Wake Forest University School of Medicine; 1999.
15 Vogler WR. An analysis of the American Cancer Society Clinical Research Training Program. J Cancer Educ. 2004;19:91–94.
16 Hirsch J. An association for patient-oriented research. Ann Intern Med. 1999;130:1014–1017.
17 Ley TJ, Rosenberg LE. The physician–scientist career pipeline in 2005: Build it, and they will come. JAMA. 2005;294:1343–1351.
20 Feinstein AR, Beck JC. Editorial: The Robert Wood Johnson Clinical Scholars Program. N Engl J Med. 1975;293:1149–1150.
21 Piccirillo JF. The Robert Wood Johnson Clinical Scholars Program: A research training plan. Otolaryngol Head Neck Surg. 1992;106:25–26.
22 Peterson AV, Fisher LD. Teaching the principles of clinical trials design and management. Biometrics. 1980;36:687–697.
23 Hulley SB, Cummings SR. Designing Clinical Research: An Epidemiologic Approach. Baltimore, Md: Williams and Wilkins; 1988.
24 Bakken LL, Lichtenstein M. Survey of the impact of National Institutes of Health clinical research curriculum awards (K30) between 1999 and 2004. J Investig Med. 2005;53:123–127.
26 Association of American Medical Colleges. Promoting Translational and Clinical Science: The Critical Role of Medical Schools and Teaching Hospitals. Report of the AAMC Task Force II on Clinical Research. Washington, DC: Association of American Medical Colleges; 2006.
27 Dickler HB, Korn D, Gabbe SG. Promoting translational and clinical science: The critical role of medical schools and teaching hospitals. PLoS Med. 2006;3:1492–1495.
28 Shuster AL, Cluff LE, Haynes MA, Hook EW, Rogers DE. An innovation in physician training: The Clinical Scholars Program. J Med Educ. 1983;58:101–111.
30 Bakken LL, Byars-Winston A, Wang MF. Viewing clinical research career development through the lens of social cognitive career theory. Adv Health Sci Educ Theory Pract. 2006;11:91–110.
31 Nabel EG. The physician–scientist: A value proposition. J Clin Invest. 2008;118:1233–1235.