THE GAP BETWEEN BASIC SCIENCES AND MEDICINE

The concept that federally funded research was in the public's best interest began during World War II when problems such as malaria, bacterial and viral infection, and trauma resulted in mobilization of the nation's scientific community. Vannevar Bush and others promoted the view that the country would benefit from federally funded research to be performed in university laboratories (National Science Foundation, 1960). However, it was the leadership and wisdom of James Shannon that resulted in postwar growth of basic science at the National Institutes of Health and of basic science departments in the nation's medical schools (Farber, 1982). In part influenced by the Flexner report (1910) the Shannon model was based on the concept that diseases will be cured only when science produces fundamental understanding of physiology and pathophysiology.

Federal funding converted U.S. universities and medical centers into research-intensive institutions. Physician-scientists and basic scientists flourished, as did scientific interactions between them. The extraordinary accomplishments of this so-called golden era of medical research have been extensively reviewed (Comroe and Dripps, 1976; Goldstein and Brown, 1997; Healy, 1988; London, 1964). Research laboratories were often built adjacent to clinical facilities to facilitate exchange between basic and clinical investigators. Becoming a physician-scientist was a highly sought goal and it was realistic to plan a career in which one could be a productive investigator, expert clinician, and outstanding teacher. Many Ph.D. scientists held joint appointments and worked collaboratively in clinical and basic science departments. The result was a homogeneous culture predicated on the premise that laboratory and bedside were interdependent as well as indissolubly linked.

By the 1970s, however, there were troubling signs. Despite the steady advances in basic science and clinical research, laboratory technologies were becoming more complex, budgets for research and training were reduced in real dollars, and a new rule (the payback provision) became a further deterrent to clinical research. In 1979, based on study of NIH grant applications for research and training, James Wyngaarden, who later became director of the NIH, was the first to express concern publicly about the declining interest in research on the part of medical students, house officers, M.D. postdoctoral fellows, and young faculty (Wyngaarden, 1979). In an article in the New England Journal of Medicine he predicted that because the academic pipeline (i.e., the time required for training prior to acquiring an academic position) was about eight years at that time, the effects he observed would not be fully manifested for another decade, which proved to be the case.

In 1984 Gordon Gill observed that physician-scientists were attracted to the power and comparatively simpler systems of molecular biology and consequently were abandoning patient-oriented research. He presciently commented, “It seems ironic that a separation of functions (i.e., science and medicine) occurred when physicians became scientists and when the work of basic scientists became clinically relevant, but such is the case, and there is no going back...the paths will not again merge” (Gill, 1984).

Indeed, over the ensuing 15 years the trend has increased and has now reached crisis levels (ASCI, 1998; Feinstein, 1999; Goldstein, 1986, 1999; Goldstein and Brown, 1997; Healy, 1988; Healy and Keyworth, 1985; Nathan, 1998; Rosenberg, 1999). Between 1992 and 1997 there was a 51 percent reduction in the number of NIH postdoctoral traineeships awarded to physicians—from 2,613 grants to 1,261 (Zemlo and Garrison, 1999). If this trend continues unabated, there will be no physicians in the postdoctoral pool by the year 2006! The irony is that physicians are not entering patient-oriented research at a time that provides the greatest opportunities for research into the cause, mechanism, prevention, and treatment of major diseases.

Many factors are responsible for the steady decline in the number of physician-scientists that contribute substantially to the gap. These factors have been discussed in several important articles (Goldstein and Brown, 1997; Healy, 1988; Healy and Keyworth, 1985; Nathan, 1998) and most recently by Leon Rosenberg in his Shannon lecture (1999). The major factors are agreed to by all who have considered the problem.

• Increased financial indebtedness of medical graduates pressures them into practice and away from the risk and uncertainties of an academic career. According to the AAMC Graduation Questionnaire of 1999, medical school graduates who were indebted had an average debt of $90,000; over 13 percent of them owed $150,000 or more, and those who attended private schools owed an average of $109,000.
• When the prolonged postgraduate training period for specialty boards is added to the time needed for training in scientific skills, 10 years may be required after medical school graduation. Fledgling physician-scientists are well into their thirties before entering the academic world.
• It is considerably more difficult today for a physician to acquire the training needed to enter a career in biomedical research than it was in the 1960s. One reason is the inadequacy of postgraduate training in medical research, as revealed by NIH outcome data (Arias, 1989; National Research Council, 1994; Nathan, 1998; Zemlo and Garrison, 1999). Few medical specialty training programs include obligatory courses in basic sciences; basic scientists infrequently serve as preceptors; and mechanisms for establishing collaboration and teamwork between basic and clinical scientists are neither identified nor widely fostered. Training in basic science demands additional time, dedication, and temporary detachment from clinical activities. Joseph Goldstein and Michael Brown suggested that M.D.s and M.D./Ph.D.s should gravitate to a career in basic science because of the seductive perception that basic science research is easier to perform successfully than is clinical research (Goldstein and Brown, 1997) and is sustained by technological breakthroughs (e.g., cDNA clones, cell lines, recombinant proteins, and monoclonal antibodies), which are readily available for application to study cellular processes in health and disease.

In reality it is increasingly difficult for an academician to be expert in each of the traditional components of the three-legged academic stool: clinical medicine, teaching, and research. With the exception of a small number of individuals who aspire to achieve this role the vast majority of physician-scientists inevitably choose one path or another but not the hybrid form.

• Increasing competition for research funding has progressively decreased the number and proportion of physician-directed research grants. Throughout a nearly 30-year interval the rates at which M.D. and Ph.D. applicants have been awarded NIH grants have been virtually identical, but physician-scientists have become a progressively smaller minority of those seeking and obtaining NIH project support (Zemlo and Garrison, 1999). The actual number of first time M.D. applicants for NIH research projects decreased by 31 percent from 1994 to 1997, without a compensatory increase in applications from M.D./Ph.D.s. Rosenberg (1999) noted that if this progression were to continue linearly, there would be no first-time M.D. applicants by the year 2003!
• Dramatic changes in the health care system, largely the advent of managed care, have shortened hospitalization periods, increased patient turnover, and de-emphasized the value of research and innovative teaching by accelerating the pace at which physicians work in a clinical setting. These changes have imposed financial constraints on all academic health centers. To accommodate for financial shortcomings clinical faculties are pressured to see more patients and earn more of their income from clinical practice, thereby accelerating a cycle that reduces research and teaching time and thus the investigator's competitiveness for acquiring research funding.
• From an educational standpoint public emphasis over the past 30 years that physicians should be directed more into primary care than into medical specialties, has resulted in changes by academic leaders in curriculum, student selection, and other clinical programs. According to Rosenberg (1999) this “ha[s] not been balanced by the equally important message that improving health of the public requires more research in which physicians must be key participants.”

All these factors have diminished the image of the physician-scientist professor as role model and thereby have contributed substantially to decreasing the number of physician-scientists and to widening the gap between advances in basic science and their application to human disease.

PROPOSALS TO BRIDGE THE GAP

Federal agencies and private foundations are considering numerous proposals to ameliorate the continued decline in number of physician-scientists, such as recruitment, debt reduction and long-range support for competitively selected physician-scientists, and training programs in clinical research, including clinical trials, epidemiology, and outcome analysis (ASCI, 1998; Goldstein, 1999; Nathan, 1998; Rosenberg, 1999; Zemlo and Garrison, 1999). Changes in the attitudes of medical school leadership and curriculum committees to enhance incorporation of advances in science into teaching and to provide research opportunities for medical students are being encouraged.

THE ROLE OF M.D./PH.D. PROGRAMS

Combined M.D./Ph.D. programs (Medical Science Training program, MSTP), begun under the auspices of the National Institute of General Medical Sciences in 1964, now involve 130 medical schools and encompass 500 students per year (Kornfeld, 1999; NIGMS, 1988). In contrast to the initial MSTP programs, which were entirely supported by NIH, the majority of MSTP students throughout the country are currently supported by private funds. Stuart Kornfeld reviewed the experiences at Johns Hopkins; Harvard; University of California, San Francisco; Chicago; Pennsylvania; Stanford; and Washington University schools of medicine (Kornfeld, 1999). In general, the results were similar in each program. The average time to graduation was 7.8 years. The duration of training was further increased because 95 percent of graduates took a clinical residency. Because MSTP students often spend four to five years in residency programs, they commonly require additional postdoctoral research training in preparation for faculty positions and competitive grant proposals. Approximately 75 percent of MSTP graduates in Kornfeld's study acquired academic positions and slightly less than 20 percent were in medical practice or industry. Almost 85 percent of graduates were engaged in research that was classified as basic and less than 10 percent were performing research classified as clinical. Although MSTP programs have proven to be successful in providing outstanding basic science-oriented physician-scientists, the training period is long (making a quick fix to the current crisis in physician-scientists unlikely) and the costs are high. In addition, given that the major objective of medical schools is to train physicians, there are limits to expanding the number of applicants who are accepted into an M.D./Ph.D. program per year. For example, currently at Washington University School of Medicine, 15 percent of all medical students are M.D./Ph.D. candidates. Because most M.D./Ph.D. graduates perform basic research, increasing their numbers, although desirable for other reasons, does not directly address the problem of increasing clinical or translational research. Thus, MSTP programs are important parts of the bridge linking basic science and human disease but cannot be considered as the sole or major component.

THE ROLE OF PH.D. SCIENTISTS IN BIOMEDICAL RESEARCH

In the late nineteenth and early twentieth centuries pathology and physiology were the dominant medical research disciplines. Major important advances were often based on clinical observations and came from chemists, some of whom were also trained as physicians. For example, Pasteur and Ehrlich attended pathology sessions and conferred with clinicians and their patients. Before World War II, medical institutions in Europe did not offer Ph.D. degrees, which accounts for the fact that the early twentieth-century leaders in biochemistry, such as Krebs, Myerhoff, Lipman, the Coris, Ochoa, and many others, were trained as physicians before becoming scientists. Increased congressional funding for biomedical science after World War II resulted in the establishment of medical school basic science departments, which increasingly produced Ph.D. graduates; in contrast to the prewar European tradition they received little or no training in pathobiology. The decline in physician-scientists began simultaneously with increased progress in basic biologic sciences. Ph.D. scientists continue to make critical contributions to the understanding of disease; however, because of the increasing pace of scientific accomplishment and the decline in physician-scientists, the gap is progressively increasing.

Many clinical investigators contributed to basic science by identifying key problems as well as by making original discoveries, and basic scientists have made discoveries that profoundly changed clinical practice. In 1964 Irving London commented on this distinction: “The essence of fundamental investigation lies not in whether it is done in a preclinical or in a clinical department or on a ward. It is rather the quality of the question which is asked and the quality of the experiment which is designed to answer the question that determines whether research is fundamental in character (London, 1964).” In those days what were limiting were more often ideas than funding or even space. Research training for physician-scientists could be accomplished in two to three years and Ph.D. scientists were relatively abundant in clinical departments where they usually held joint appointments in basic science departments. Even medical grand rounds were frequently shared between clinical and basic scientists who discussed patients and their illnesses.

All components of this seemingly idyllic existence have changed. Since the 1970s the gap between basic science and medicine has increased largely because science has become more complicated. Clinical scientists have greater difficulty in applying these advances to disease and basic scientists are needed. Many problems are so complex they exceed the ability of traditional clinical scientists to deal with them; others are less complex but necessitate collaboration between basic and clinical scientists. Unfortunately most basic scientists have little knowledge of pathobiology or clinical medicine; therefore, it is logical that basic scientists should learn enough pathobiology to attack disease-related problems in collaboration with physician-scientists. Ph.D. scientists cannot replace physician-scientists in performing clinical or translational research.

The goal is to have Ph.D. graduates who can function in the interface between basic science and disease and collaborate with physician-investigators who work in the interface between the patient and basic science. Achieving this objective is made more difficult because training in basic science is usually absent from the third and fourth year medical school curriculum and from postgraduate residency programs.

TRAINING PH.D.s IN PATHOBIOLOGY

As in a structural bridge there are many components to bridging the gap between the advances in basic science and disease. Virtually all efforts to bridge the gap have been based on the premise that “biomedical research is tightly linked to physician manpower” (Healy, 1988; Healy and Keyworth, 1985; Rosenberg, 1999; Wyngaarden, 1979). Most notable are M.D./Ph.D. programs and public and foundation efforts that are directed at students at every academic level. One approach that has not received much attention concerns the role of Ph.D. students, fellows, and graduates.

Almost every basic science department in our medical schools has a graduate program. Based on an ongoing poll of 372 graduate students in two leading institutions, 97 percent of the students chose training in a medical school rather than in a university because they sought careers that impact on human health. More often than not their graduate training was highly focused and not different from what could have been obtained at a university. A survey of 98 Ph.D. graduates who spent over six years in great medical centers revealed that the majority of the graduates had little knowledge of basic mechanisms of pathology, how their research related to organ physiology or pathophysiology, or what constitutes research in a clinical setting. In addition, their career directions were virtually the same as those of students who received a Ph.D. from university basic science departments. More than 90 percent of students graduating from medical school basic science programs sought the same goals as did graduates of university-based graduate programs in biology or chemistry, namely, stable positions in basic science departments, research institutes, or industry. Interest in pathobiologic mechanisms had severely waned and knowledge of disease processes, including pathology, diagnostics, and therapeutics, was deficient. It is not uncommon for such graduates, despite their research brilliance and ability, to be unable to describe what, for example, inflammation, necrosis, or fibrosis look like, let alone what they may feel like to a patient. The reason is that few graduate programs teach pathobiology, and many thesis advisers in basic science departments have little knowledge and interest in disease mechanisms or clinical problems.

These students are unaware of the changing scene in academic clinical departments and the increasing opportunities for Ph.D.s graduates to have productive careers in clinical departments. Longstanding academic problems regarding the role of a Ph.D. scientist in a clinical department are slowly changing primarily as a result of the decline in physician-scientists. In many institutions Ph.D. graduates are not attracted to a primary academic appointment in a clinical department because their scientific independence and academic tenure are limited or nonexistent. One of my students succinctly described the situation: “You work for and not with a physician.” As research funding for physician-scientists declines and advances in basic science continue, medical centers are under increasing pressure to restore research efforts and solve the academic problems associated with recruitment of basic scientists into clinical departments. As will be apparent in data to be presented later, a 1989 proposed scenario has proven to be at least partially correct: “According to this scenario basic scientists who have been trained in pathobiology will have an exciting opportunity for productive careers in clinical departments” (Arias, 1989).

Literature search reveals two brief published letters but no detailed commentaries on the teaching of Ph.D. scientists in biomedical research before the 1970s. The major reason is that before the increasing gap the problem, if it existed, was not of major concern. There were some exceptions. Irving London's 1964 presidential address to the American Society for Clinical Investigation described the excitement of basic research, its importance to medicine, the reciprocal and interactive relationship between basic science and medicine, and anticipated some problems that could result from the increasing complexity of science (London, 1964).

In 1979 Morris Karnovsky observed that graduate students at Harvard, although outstanding in their knowledge of basic science, knew comparatively little about organismal physiology and disease mechanisms (M. Karnovsky, personal communication, course in pathology and pathophysiology at Harvard Medical School, 1980). With support from the Josiah Macy Foundation, Karnovsky created a one-semester course that during the subsequent five years accommodated approximately 50 graduate students from institutions in the greater Boston area. The course involved lectures on histology, the general basis of pathology, and reactions to injury in major diseases. Gross and microscopic specimens were frequently demonstrated. Participants reviewed and discussed original basic and clinical research papers. The course was successful and continued for five years, at which time funding ceased. Further support was not forthcoming from other foundations, industry, or the university. No formal outcome studies were performed, but Karnovsky recalls that attendance was full, enthusiasm was high, and many students wrote that the course changed their career interest to pathophysiology (personal communication). This novel course tapped into the unfulfilled interests of Ph.D. graduate students in a medical center. Although these interests have not diminished with time, I am unaware of other similar academic ventures from the late 1970s until 1984.

In a 1989 article in the New England Journal of Medicine I proposed that Ph.D. students, postdoctoral fellows, and faculty receive training in pathobiology as part of the effort to bridge the gap between basic science and its application to medicine (Arias, 1989). A one-semester course in pathobiology for Ph.D. students, fellows, and faculty was described. Colloquially it may be said that the goal of the course was to demystify medicine for Ph.D. students, fellows, and faculty members. There are several unique features to the course.

• Participants see patients, handle pathologic specimens, and are exposed to most of the major diagnostic and therapeutic facility in a modern hospital.
• Clinical and pathology sessions regarding approximately 20 major diseases are followed by Socratic-style analysis of the related basic biologic problem (e.g., growth control, autoimmunity).
• Students are given substantial reading material for analysis, but the course is intended to elicit their enthusiasm, long-term interest in pathophysiology, and understanding of where basic science and disease intersect.
• The course is available to all graduate students, fellows, and Ph.D. faculty at Tufts and is heavily oversubscribed. It is limited to 15 individuals at a time when a larger group would make the clinical activities impossible. The group invariably consists of seven or eight second- to sixth-year graduate students in different graduate programs, three to five postdoctoral fellows mainly from an NIDDK Training Grant in Molecular and Cellular Pathophysiology, and one to four basic science faculty, visiting faculty from institutions seeking to replicate the course, or biotechnology or pharmaceutical company scientists. Heterogeneity in the group has been important to group dynamics and learning.

Outcome data are available because the course has been given for 15 years and we have followed all participants. The results are encouraging and unambiguously support the desirability of such activities. From 1984 to 1998 there were 214 participants in the course; 151 were graduate students, 42 were postdoctoral fellows, 13 were Ph.D. basic science faculty, and 8 were biotechnology and pharmaceutical industry scientists. By 1998, 88 individuals had completed all postdoctoral training and entered the academic arena. Twenty-seven (30 percent) have tenure-track positions in basic science departments, mainly in medical schools; 38 (40 percent) have responsible pathobiology-oriented positions in leading biotechnology and pharmaceutical companies, and 23 (25 percent) have tenured track positions in medical school clinical departments throughout the country. The departments and the distribution of graduates include medicine (15), parasitology (2), pathology (4), neurology (1), and pediatrics (1). The represented divisions in departments of medicine and the number of graduates are endocrinology (2), cardiology (1), gastroenterology (3), pulmonary (2), hematology-oncology (3), infectious disease (2), and immunology (2). Only 3 graduates who completed postdoctoral training are not currently working in science, 2 of which are recent mothers! Each of the 8 biotechnology scientists who participated in the program directs a major human disease research effort. Six Ph.D. faculty who participated in the program subsequently acquired NIH grants in collaboration with physician-scientists in clinical departments.

The course has been funded progressively by the Josiah Macy Foundation (1984-1988), Lucille P. Markey Charitable Trust (1989-1996), a Boston-based private foundation (1997-1999), various private donors, and a NIDDK Training Grant in molecular and cellular pathophysiology. Aside from stipends provided by the training grant, the annual cost is $ 60,000, which is largely spent for supplemental support for graduate students, supplies, printing, and administrative assistance. Nineteen institutions in the United States, Canada, and Europe have expressed interest in our program and 11 have begun similar activities.

Karnovsky's course at Harvard and our experience at Tufts played a part in influencing the leadership of the Lucille P. Markey Charitable Trust to support additional programs in pathobiology. From 1992 to 1996 the Markey Trust supported eight other programs that specifically seek to bridge basic science and medicine by instructing basic scientists in pathobiology. As described in this volume (Bunn and Casey, 1995; Lucille P. Markey Charitable Trust, 1996), each program has a different orientation and composition. Most involve additional time in graduate school, some provide degrees (M.S. or Ph.D.), all provide student stipends and some include clinical demonstrations. The Tuft's program is unique in that the duration of graduate school training is not increased, gross and microscopic pathology are presented, and selected patients are seen in a clinical setting. Participants also become acquainted with most major diagnostic and therapeutic facility in a modern medical center.

Unfortunately funding by the Markey Trust ended in 1998, and there is little outcome data on graduates of the other eight programs. Of greater concern is that despite the demonstrated need for and interest in demystifying medicine for many Ph.D. scientists, no other major funding source has assumed the mantle for sustaining and encouraging further development of this important component in our academic bridge or for performing outcome studies of existing programs. Whereas outcome evaluation of basic and clinical scientific studies is required for their continuing support, outcome evaluation of educational and training programs invariably lacks support from government agencies and private sources.

Several other graduate programs, such as Edward Kravitz's course at Harvard Medical School on the pathobiology of neurologic disease (E. A. Kravitz, personal communication, 1989), have created disease-oriented courses for their students and fellows. The general format involves lectures, which are often supplemented by presentation of patients and discussion of their illnesses. The goal of these programs has not been to direct Ph.D. scientists into clinical studies but to demonstrate how the fundamental research they are engaged in is relevant to human disease.

It is hoped that with increased awareness of the need to rejuvenate clinical investigation, there will be accompanying efforts to benefit from recent experiences regarding the training of Ph.D. scientists in pathobiology. Our experience does not indicate that Ph.D. scientists can replace physician-scientists in the study of human disease, particularly at the clinical level. However, Ph.D. graduates are important struts in the bridge that links basic science and medicine, and their incorporation into the process seems timely and long overdue.

CONCLUSIONS

Bridging the increasing gap between advances in basic science and their application to medicine requires changes in the training of both basic scientists and clinical investigators. The following conclusions regarding the education and training of physicians are based on a personal assessment of available published data and commentaries by others:

• Premedical students should be advised that quantitative skills in biology, chemistry, and mathematics are increasingly required for careers in research or in medical practice.
• Medical school teaching should be restructured to make research opportunities more readily available for students and residents and provide teaching of scientific advances throughout the curriculum. Greater selectivity in courses should be provided for medical students who are seriously interested in research.
• Reconsider the timing, duration, and number of M.D./Ph.D. programs to make them more efficient and productive.
• Revitalize physician-scientist recruitment and training by addressing the problems that contribute to the gap.

The following conclusion regarding the training of Ph.D. students and scientists is based on our outcome data:

Encourage and support programs to train Ph.D. students and fellows in pathobiology. Ph.D. students in medical school graduate programs seek careers that bridge with human disease, however, most students graduate with little knowledge of human clinical disease or pathology. Changes in medical research and the decline in physician-scientists create exciting opportunities for Ph.D. graduates to work with but not for physician-scientists as tenure track faculty in clinical departments. The bridge between advances in biological science and medicine has many components, including pathobiologically versed Ph.D. scientists who supplement but do not replace physician-scientists or clinical investigators. Pathobiology programs for Ph.D. students and fellows meet student and society's needs, are a good investment, and should be encouraged and supported nationally.

Bridge building in biomedical research has parallels with a structural bridge, which serves as a useful metaphor (Shapiro, 1983). Both require many different kinds of parts, each of which is essential for proper function. Bridge traffic is bidirectional. Once it has been erected, life on either side of the bridge is no longer as it was. The challenges inherent in both bridge building and maintaining its integrity transcend the merely structural. Meeting those challenges requires imagination, dedication, creativity, and a willingness to take risks.

REFERENCES

• Arias, I. M. 1989. Training basic scientists to bridge the gap between basic science and its application to human disease. New England Journal of Medicine 321:972-974. [PubMed: 2779621]
• American Society for Clinical Investigation. 1998. Position paper for Proceedings of the Institute of Medicine Conference on the NIH Research Priority Setting Process. Washington, D.C. Available at <http://www​.asci-jci.org/asci/>.
• Bunn, H. F., and C. G. Casey. 1995. Educating the biomedical scientist. FASEB Journal 9:1392-1395. [PubMed: 7589979]
• Comroe, J. H., and R. D. Dripps. 1976. Scientific basis for the support of biomedical science. Science 192:105-111. [PubMed: 769161]
• Farber, S. J. 1982. Presentation of George M. Kober Medal to James A Shannon. Transactions of the Association of American Physicians 95:cxxix-cxlix. [PubMed: 6763808]
• Feinstein, A. R. 1999. Basic biomedical science and the destruction of the pathophysiologic bridge from bench to bedside. American Journal of Medicine 107:452-457. [PubMed: 10569301]
• Flexner, A. 1910. Medical Education in the United States and Canada . New York: Carnegie Foundation.
• Gill, G. 1984. The end of the physician-scientist? American Scholar 53(3):353-369.
• Goldstein, J. 1986. On the origin and prevention of PAIDS (paralyzed academic investigator's disease syndrome). Journal of Clinical Investigation 78:848-854. [PMC free article: PMC423687] [PubMed: 3528221]
• Goldstein, J. L. 1999. Congressional Testimony to the House Subcommittee on Labor, HHS, and Educational Appropriations, April 28. Daily Digest, p. D450, Available at: <http://thomas​.loc.gov​/cgi-bin/query/D?r106:27:​./temp/~r106WQfvjc>.
• Goldstein, J. L., and M. S. Brown. 1997. The clinical investigator: Bewitched, bothered and bewildered—but still beloved. Journal of Clinical Investigation 99:2803-2812. [PMC free article: PMC508127] [PubMed: 9185499]
• Healy, B. 1988. Innovators for the 21st century: Will we face a crisis in biomedical-research brainpower? New England Journal of Medicine 319:1058-1064. [PubMed: 3173434]
• Healy, B., and G. A. Keyworth. 1985. The NIH and numbers: A vital concern's concerns. New England Journal of Medicine 312:1450-1452. [PubMed: 3990746]
• Kornfeld, S. 1999. M.D./Ph.D. programs: Are they successful and can they fill the gap? Physicians-scientists and career opportunities for biomedical research. Presented at the Federation of American Societies for Experimental Biology Annual Conference, Washington, D.C.
• London, I. M. 1964. The impact of the revolution in biology on clinical investigation. Journal of Clinical Investigation 43:1222-1224. [PMC free article: PMC289602] [PubMed: 14171797]
• Lucille P. Markey Charitable Trust. 1996. Final Report 1983-1996 . Miami, FL: Lucille P. Markey Charitable Trust.
• Nathan, D. G., for the National Institutes of Health Director's Panel on Clinical Research. 1998. Clinical research: perceptions, reality and proposed solutions. Journal of the American Medical Association 280:1427-1431. [PubMed: 9801003]
• National Institute of General Medical Sciences. 1988. The careers and professional activities of graduates of the NIGMS medical scientist training program. MSTP study. Available at <http://www​.nih.gov/nigms​/news/reports/mstpstudy/mstp-prinet​.html>.
• National Research Council. 1994. Meeting the Nation's Needs for Biomedical and Behavioral Scientists. Appendix C, Public hearing on national needs for biomedical and behavioral research personnel. Washington, D.C.: National Academy Press.
• National Science Foundation. 1960. Science, the Endless Frontier. A Report To The President on a Program for Postwar Scientific Research. Washington, D.C.: U.S. Government Printing Office. [PubMed: 17813931]
• Rosenberg, L. 1999. Physician-scientists-endangered and essential. Science 283:331-332. [PubMed: 9925491]
• Shapiro, M. J. 1983. A Picture History of the Brooklyn Bridge . New York: Dover Publications.
• Wyngaarden, J. B. 1979. The clinical investigator as an endangered species. New England Journal of Medicine 301:1254-1259. [PubMed: 503128]
• Zemlo, T., and H. Garrison. 1999. Is there a shortage of physician-scientists? Physician-scientists and career opportunities for biomedical research. Presented at the Federation of American Societies for Experimental Biology Annual Conference Washington, D.C.: 1999.

Footnotes

1

The author wishes to thank many colleagues, particularly Samuel Silverstein, Ezra Lamdin, and Lyuba Varticovski, who provided valuable information, discussion, and critical review.