Basic science research increased at an unprecedented rate in the last 50 years as a result of many factors, the two most important being the American research university supported by the U.S. government and the pharmaceutical industry, which learned quickly the value of highly focused research directed at human disease and the control of certain physiologic functions. The federal expenditure for medical research in 1940 was $3 million. The total national expenditure from all sources that year was $45 million [1]. In 1990, federal support for health-related research was $11.3 billion, of which the largest component was from the National Institutes of Health (NIH), the budget for which was slightly more than $7 billion; and total national financial support was $22.5 billion, representing nearly a 500-fold or 50 000% increase in 50 years [2]. These figures would indicate superficially that health research has never been better. However, the physician who wants to participate actively in improving the human condition faces many difficulties. For this essay, the clinical investigator is the physician who may be doing basic research or clinical, patient-related research.

Historical Perspective

Clinical research is a new discipline. Before the mid-1950s, advances in patient diagnosis, treatment, and care were based on imprecise and often biased observations. The first randomized clinical trial, which was an evaluation of streptomycin in tuberculosis, was published by the British in 1948 [3]. The first randomized clinical study in the United States was reported in 1951; it was a study designed to determine the effectiveness of penicillin in pneumococcal pneumonia [4]. Researchers realized during this time that a large number of patients entered into a clinical study allowed them to draw their conclusions more quickly and increased the chance that the findings would be important. Furthermore, investigators soon appreciated the power of statistical analysis, which permitted the evaluation of large groups to determine, with some degree of certainty, small differences among treatments. Soon clinical research was accepted as a legitimate academic discipline, and two large national cooperative clinical research groups were formed. The Acute Leukemia Group published the results of the first randomized treatment trial for leukemia in 1958 [5]. In 1960, the Eastern Cooperative Oncology Group reported their randomized study of the treatment of adults with solid tumors [6]. Elegant clinical studies had been conducted previously, such as that by Shannon and colleagues [7] for the treatment of malaria, but the rigorous scientific approach required to determine the validity of small differences and to simultaneously evaluate multiple variables had never been appreciated by clinical investigators.

This was an exciting time in the development of clinical investigation. It offered a rigorous methodology to extend laboratory discoveries to humans. The nation was proud of these collective achievements in medicine. The number of medical schools increased from 77 to the current 126. The Hill-Burton Act allowed use of federal funds to help redevelop our hospital infrastructure, a necessity for the expansion of the nation's health initiative. Although there was no federally established and articulated goal, such as that of the space program, there was an understanding that the United States was devoted to achieving better health. The rapid development of the effective polio vaccine, introduced by Salk in 1955 [8], solidified the public's belief that enhanced health care and health research were important. This was an answer to a growing national concern about health and helplessness. The consequence of these and other successful discoveries was the creation of a new industry, the health services and research industry. Now our desire to pay the costs related to health care may be saturated, but we continue to demand better health care and more health research. We now face a monetary crisis and a crisis in which too few physicians are willing to make research in health a career. Furthermore, this crisis might be ignored if the mandate to produce more primary care physicians is unilaterally accepted without considering the effects this will have on other equally important health-related goals of our country.

Research Funding

It is a fact that physicians on medical school faculties are having increasing difficulty in obtaining funding to support their research and therefore themselves. The money to support medical school faculty comes from four primary sources: the NIH, private foundations, private industry, and clinical activity. The chance of receiving funding from the NIH for a new investigator-initiated application (R01) in 1990 was 14.2%, even though the application was considered worthy of being done (that is, it was not disapproved). This is considerably less than the 40% success rate in 1965 and 1975 [2]. If the application was a renewal of a previously funded grant, the success rate in 1990 was 35% [2], which is in marked contrast to the nearly 70% funding rate for renewal grants submitted in 1965 and 1975 (Figure 1).

View larger version (34K): [in this window] [in a new window]   Figure 1. New and renewal investigator-initiated research grant applications (R01) to the National Institutes of Health (U.S. Dept. of Health and Human Services). By outcome: disapproved, unfunded, and funded from 1965 through 1990 [2].

The number of NIH grants funded in each year during the last decade has not changed substantially. In 1990, of the 25 725 total, 15 888 were R01s, of which 4845 were newly funded R01 grants, which includes both new and renewal applications, and 2083 were first awards [2]. In 1991, for which the most recent data are available, 2225 new R01s and 449 new first awards were made [9]. The new grant is less likely to be funded than is the continuation of an ongoing NIH-funded project. This is discouraging information to the new clinical investigator. A few grants have gone to program grants (771 in 1990), but the funds appropriated for this mechanism clearly do not substitute for the current reduced funding rate to new applicants. Nearly one third (9066) of all grants are not R01s but rather are "other" grants, which includes center grants, special projects, and disease-oriented requests for proposals (Figure 2).

View larger version (27K): [in this window] [in a new window]   Figure 2. Grants funded by the National Institutes of Health. By type of grant: investigator-initiated (R01). Program Project Grant (PO1), and others from 1981 through 1990. In 1990 there were 25 725 total, 15 888 R01s, 9066 other, and 771 PO1s. Only 4845 of the R01s were newly funded in 1990.

The percentage of new R01 applications submitted by physicians, including those with both MD and PhD degrees, has remained nearly 25% of the total since 1980 [10]. The priority scores of MD applications are nearly identical to the applications made by researchers with a PhD degree, which shows no particular preference for funding based on the degree. However, the assumption that only 14% of new R01 applications were funded in 1990 would indicate that only 386 new applications of the 2762 submitted from physicians were funded in that year. The information is not available, but it would be reasonable to assume that many of these new applications were from established investigators and not from new physicians just beginning an investigative career.

In 1991, there were nearly 41 000 full-time faculty of clinical departments in U.S. medical schools [11, 12], of which 12 892 were faculty of departments of internal medicine (or 102 per medical school) [12, 13]. Faculty with both MD and PhD degrees in internal medicine departments numbered only 954. If 50% of the 386 new grant applications funded by the NIH were from new faculty, 193 physicians would have felt their decision to begin an academic career in clinical investigation was wise. I wonder what the other 1400 young physician-scientists whose applications were denied funding by the NIH considered in 1990? If these young faculty wish to pursue their dream of clinical investigation in an academic setting, they have only a few choices available.

The research proposal can be resubmitted with adherence to the recommendations of the reviewers. The percentage of amended grants resubmitted in the new R01 pool has increased from 16.4% in 1980 to 34.4% in 1990; for competing applications, the increase has gone from 17.5% in 1980 to 41% in 1990 [10] (Figure 3).

View larger version (26K): [in this window] [in a new window]   Figure 3. Outcome of resubmitted and amended investigator-initiated grant applications (R01s) that were previously funded from 1980 through 1990 [2] .

Of concern is the fact that 56% of amended renewal R01 applications are funded compared with the success rate of 35% for unamended renewal applications of a currently funded grant, and the 14% success rate for the initially submitted new R01 grant. The perception that a resubmission of a renewal grant after an initial nonfunding cycle is likely to be funded is clearly a valid assumption. As a result, researchers commonly apply one cycle early so the amended resubmission can be submitted so that funding will not be interrupted. This environment is not encouraging to new investigators. How could an application be so fundamentally flawed that funding is not provided, yet 6 months later the same idea and premise, with some attention to a reviewer's comments, becomes acceptable and funding is provided?

We are creating a culture that accepts the concept that multiple resubmissions are required to receive funding. Using a recent estimate, an investigator must submit six grant applications to get one funded. Moreover, with an average grant duration of 4.1 years [2], an investigator who wishes to have stable funding from the NIH must write a grant application every 8 months [14]. These projections assume that 90% of all grants are worthy of funding. Clearly, bad ideas will probably not be funded, regardless of the number of submissions. Reviewers could help investigators by disapproving poor or marginal grants. The nearly nonexistent disapproval rate of a few percentage points promotes resubmissions. This system encourages and rewards reapplication and is one many young clinical investigators will tolerate only for a few years. Some physicians now refuse to begin an academic career they so much coveted only a few years earlier because of the politics of funding.

There are alternatives to the R01 funding mechanism. The young clinical investigator can become a part of an established investigator's program and obtain funding through a large R01 of the senior faculty member or join a larger program project grant. The complexity of many research problems demands that larger laboratories or collaborative research programs be developed. In this environment, new investigators will have opportunities to participate and even become well trained, but these alternatives may inhibit what is most appealing to many young faculty: the creativity expressed through the design and completion of their own ideas.

Young faculty members can appeal to private foundations, such as the American Heart Association, the American Cancer Society, the National Kidney Foundation, and others to support their projects, but these avenues are no less difficult than the NIH.

Private industry may be a source of funds for certain projects, especially if corporate research plans include the faculty member's field. However, this research may be so directed by the company that the investigator's creativity is discouraged or even prevented. A recent survey of young internal medicine faculty revealed that 40% had received financial support from a manufacturer at some time during their tenure [15], which indicates the extent to which private industry contributes to medical school faculty. However, these funds may soon be greatly reduced because of constraints the Clinton health care plan will have on the research and development budgets of the pharmaceutical and biotechnology industries.

Adopting the entrepreneurial spirit has become acceptable in the last decade. If an idea is sufficiently clever and may influence human well-being in the near future, venture capital may provide the necessary money to pursue the investigation for several years. These latter two situations, accepting financial support from private industry or from venture capital groups, or from both, to establish an investigator-directed new company, have caused considerable debate and prompted the development of guidelines to help investigators avoid conflicts of interest [16-19].

Finally, the clinical investigator can devote more time to clinical care, which is what presumably attracted him or her to medicine. Patients generally appreciate their physician's work, which contrasts with the feeling the young clinical investigator experiences when the grant that took 6 weeks to write has been relegated to approved-but-unfunded status. In addition, there is more certainty in financial remuneration when patient care is performed. The more patients one sees, the more money one earns. In contrast, a reliable association between academic research and financial security no longer exists. Personal satisfaction from writing many unfunded applications dissipates quickly. Encouragement from a senior faculty member or a department chair means less and less after each research rejection.

Most faculty are responsible for different fractions of their salary. Some departments provide "start-up" funds for 3 years, a period during which the clinical investigator is expected to become financially self-sufficient through any of, or a combination of, the mechanisms mentioned above. The clinical investigator's concerns can be somewhat assuaged by knowing that patient care can be increased to fill the time reserved for research for which there is no money, and anxiety about salary support can also be reduced. However, the investigator with only the PhD degree does not have this "safe haven" when funding becomes problematic and thus has few alternatives but to change careers or join the private research industry [20].

Investigators will go where the funding is. To support research from clinical income is not practical for most faculty for one reason: Physicians practicing internal medicine must devote 100% of their time to patient care to earn enough to pay their salaries. (This might not be so for some surgery faculty, but information is not available for appropriate comparison.) Young physician-investigators quickly learn this after a few years of gradually increasing the percentage of their time in patient care. Therefore, physician-investigators have limited choices. To leave the full-time faculty has some advantages, with the major incentive being a reduction in expenses, which include department and school taxes not required of the private-practice physicians.

Physician-investigators can also search for research funds from other sources. As discussed earlier, the growth in health-related research and development funding has been greatest in private industry [2] (Figure 4). Sometimes faculty members suddenly find that private companies offer them the opportunity to join their group with job and financial guarantees that are unmatched by medical schools.

View larger version (29K): [in this window] [in a new window]   Figure 4. National financial support for research and development in health from 1981 through 1990. Categorized by private nonprofit, industry, state and local governments, non-National Institutes of Health federal, and National Institutes of Health [2]. NIH = National Institutes of Health.

Work Environment

The physician-investigator must balance many responsibilities, including research, teaching, and caring for patients. It is difficult, perhaps impossible, to do all of these three traditional academic activities well. The competition for research funding is so fierce that the likelihood of obtaining financial support presumably would be enhanced by devoting greater effort to research. However, conducting research to the exclusion of other activities may not guarantee funding. In 1986, a study of all internal medicine faculty conducted by the Association of Professors of Medicine, in conjunction with the Association of American Medical Colleges, found that the median research effort of physician faculty was 25%, whereas faculty with PhD degrees had a median effort of 95%. Only 23% of the physician faculty were principal investigators of an NIH grant (that is, R01); however, 67.9% had some form of external funding to support their research efforts, and only 25% had no research funding [21]. In 1989, a follow-up study showed that there was no difference among physician faculty (59.6%), those with both MD and PhD degrees (61.3%), or faculty with only the PhD degree (57.6%), who had ever been or were a principal investigator of a peer-reviewed research application [13]. This information, when considered in concert with the fact that the success rate of NIH R01 applications from physician and PhD investigators is nearly identical, would indicate that for some physician faculty, spending more time in research may not result in a dramatically increased number of funded grant applications. A project to evaluate percentage effort with the rate of peer-reviewed funding would provide interesting information.

To assume blindly that the less a faculty member does for the department in such areas as reduced teaching, committee work, attending, and patient care, the more productive that person's research will become is incorrect. Some young faculty, unfortunately, believe that the less they do for the department, the better their long-term prospects for research. The numerical representation of percentage effort as a reflection of actual hours spent in a certain area of work may not adequately represent the quality of that effort. In 1985, the Stanford Department of Internal Medicine showed that their faculty worked an average of 60 hours each week. The mean percentage of research time was 21.1%, and the mean percentage of time spent in patient care was 30.1% [22]. The annual gross revenue per faculty member from research activity was $944 000 and from the faculty practice plan, $249 750 (for each cardiology faculty member the figure was $507 000). The research effort expended by the Stanford faculty was clearly the most economically productive fraction of the faculty's time. However, to assume that greater time in research would have led to more research funding is not appropriate.

A third study of 482 physicians who were first authors of a paper published in one of 18 peer-reviewed journals during a 6-month period in 1986 revealed that 30% of their working time was devoted to clinical activities and 36% was spent doing research [23]. Therefore, a reasonable estimate of a current faculty's effort is one third in research, one third in clinical activity, and one third in teaching, administration, and reading.

A faculty member in a department of medicine has certain responsibilities that are different from those of faculty in research institutes, private industry, or basic science departments. Clinical faculty must collectively provide care for a group of patients. In addition, it is the department's obligation to teach clinical medicine to students, housestaff, and clinical fellows. Rather than expect all faculty to contribute equally in all areas, a division of clinical faculty and research faculty is being made. This structure is meant to provide the selected clinical investigator more time to do research. However, as some faculty assume a greater burden of the patient care and teaching activity, others get "protected time," which allows them to see fewer patients, teach less often, and serve on fewer committees. This concept of an emerging division among medical school faculty into research and clinical faculty was described by Robert Petersdorf, then Dean at the University of California, San Diego, as perhaps the future of internal medicine departments [24]. The economic necessity to maintain a department in financial balance has placed a greater emphasis on the clinical care component of a department's activities. Although it is noble to care for patients, it is our duty, as medical faculty members, to improve human health and well-being through research.

If we do need more clinical faculty, it is an opportunity to have a greater effect in patient-directed clinical investigation. If the departments of internal medicine do not ask and answer the critical clinical questions, who will? Certainly not those engaged in full-time private practice or those in full-time laboratory research. Who better to identify the problem and ask the question than the physician-investigator caring for the patient? Separation of faculty into research and clinical groups has been accomplished in some schools of medicine [25, 26].

As medical school departments of medicine have changed, the new clinical investigator has too often been forgotten. Some new faculty have been hired without clear plans described by their chair, or section chief, for their career development. The difference between satisfied and dissatisfied young faculty is not desire, training experience, "protected time," salary, or clinical work load. Rather, it is a lack of mentorship. Many new faculty share offices with others and share research space with their senior faculty advisor. However, when new faculty members are counseled on a regular basis and allowed to begin to publish on their own, they feel they are progressing and their prospects for success in a competitive environment are growing.

When asked, internal medicine faculty responded that having a mentor most influenced their choice to be a physician-investigator [13]. A rigorous evaluation of the components of a successful academic career confirmed the value of a mentor [27, 28]. This study showed that supportive and guiding mentorship during the first few years of a faculty member's experience was critical for launching a productive career. In addition to providing ongoing supervision and training and showing the new faculty member the mechanics of being a successful investigator, the most respected mentors offered emotional support. The true mentor provides assistance and advice, without the expectation of being included in the published accomplishments of the young investigator. A mentor knows when to encourage independence and derives personal satisfaction from helping another young investigator become someone else's mentor. In a survey of 15 academic leaders in general internal medicine done by Eisenberg [29], a supportive and stimulating mentor was thought to be important for development of research faculty in general internal medicine.

Financial Rearrangement at the Medical School

Professional income of medical schools exceeded research income in 1985 and has progressed at a rate that documents the emphases placed on patient care by full-time faculty [30]. In 1990, 29.8% of the revenues of all U.S. medical schools came from clinical practice, whereas in 1980 it was 15.6%. The percentage of revenue from research activity was 17% in 1990 and 20.8% in 1980. In the 1960s, consistently 40% of a medical school's revenues were derived from research, and 30% to 35% of all revenues were from the federal government to support research. By 1990, only 13% of the medical school's revenue was earned from the government for research activity [12]. Certainly, research funding was easier to obtain in the 1960s, with a greater fraction of total budget from government for research and a funding rate of 40% for new R01s and a 70% rate for renewal applications.

As medical schools grew and their clinical responsibilities expanded, so did the need to generate clinical revenue to support them. In conjunction with this clinical expansion, research funding became a smaller part of the budget. Consequently, the pressure to generate clinical income to support the academic and research activities of the faculty became even greater. This situation has had a substantial effect on the research mission of many departments of medicine.

Training

The results of several studies assessing the effect of training on a research career were the same: The better the training, the better the research outcome. A survey conducted by the Association of American Medical Colleges and reported by Levey and associates [31] concluded that the "successful researcher in academic medicine should have 2 or more years of postdoctoral research training, including formal course work in the fundamental sciences pertinent to biomedical research; 2 to 3 years of full research support from the academic institution until the first extramural grant is obtained; and commitment of at least 33% of time to research activities." A larger survey of 12 000 faculty confirmed the relationship between research training and success as an investigator [13].

The same concern that has been expressed about the training of the clinical investigator who is focused on a laboratory career has not been uniformly expressed by the leaders of medicine for those who want to become clinical investigators with patient- or health-related research as their career goals. Editorials that question the future of biomedical research emphasize the need for more funding and greater efforts to encourage the development of the physician-scientist [32-35]. Kelly [36] presented a spectrum of faculty research activity that placed the investigator devoted to clinical research in proper perspective with the basic science investigator.

Only one program has been solely devoted to the training of new young physicians who want to become clinical scholars. Since 1973, the Robert Wood Johnson Foundation has funded six national Clinical Scholars Programs [37]. This program provides future faculty with 2 years of rigorous training in the methods of clinical investigation. Other private foundations support young faculty in the early phase of their career development, including the Kaiser Foundation and the Kellogg Foundation.

Because of the nearly universal lack of support for patient-related research training, few faculty appreciate the importance of patient research. Clinical faculty are generally hired to care for patients and teach, with scholarly contributions assuming a secondary reason for the recruitment. New clinical faculty have few, if any, senior faculty who can serve as appropriate mentors and advise and guide their clinical research development in the same way that research mentors guide new laboratory investigators. Consequently, the clinical research effort often becomes case reports, descriptive reviews of patients, and historical reviews frequently prompted by the clinician's recent observation of a patient with an unusual clinical problem. This is not clinical research but rather descriptive cataloging and archiving. Modern clinical research is as sophisticated as the methods to clone genes, but few basic scientists believe this. Some senior clinical faculty, who consider clinical research the equivalent of assigning patients a random number and treating them with one of two treatments, do not believe this either.

Because of limited formal training in the methods of patient-related research for the physician-investigator, good, energetic new faculty are often ill prepared to conduct scientifically credible clinical research. Another difficulty is the prevailing attitude that patient-related research is easy and can be accomplished at night or on weekends after spending 60 hours caring for patients and teaching.

As noted, at least 2 years of postdoctoral training appears to be necessary for the most successful faculty when judged by NIH funding. It is not known what duration of training is optimal, but it should depend on the amount and type of information taught and the extent that practical experience is part of the training. Training support is not a substantial component of the NIH budget; it is 4% of the total. Nearly all of the training support goes to laboratory training, in contrast to clinical research (patient-related) training [2]. However, 57.6% of research training of physicians is funded by the NIH [13]. Of the 12 497 trainees sponsored by the NIH in 1990, only 2798 were physicians [2] (Figure 5).

View larger version (34K): [in this window] [in a new window]   Figure 5. Training grants provided by the National Institutes of Health. By recipient, predoctoral students, PhDs, and MDs from 1981 through 1990 [2].

Training grants from the NIH prevent funding during the clinical component of the physician's training, regardless of the importance that clinical training might have on the subsequent research development of the prospective faculty member. This national attitude only perpetuates the notion among laboratory investigators that patient care is not an important component of research training and patient care should be able to generate the funds to support any research, or, in this case, training. What do young clinical faculty think? Many think that to become "real" faculty members they must do laboratory research, even if they really want to do patient-related research. In the survey of internal medicine faculty, 52.2% who did research did laboratory research, whereas only 7.4% did patient-related research and 28.6% were involved in both [13]. These results would indicate that no suitable mentors exist for many students and residents when so few clinical investigators devoted to patient-related research serve as faculty.

The perceptions of the clinical investigator by those who do basic research are justified in that most clinical investigators are not trained in the scientific methods of clinical research. Of medical research faculty, 55% stated they received no formal course work during their training, and only 1.7% of the total training time of the remaining 45% of faculty was devoted to required courses [13]. The result of this poorly directed training could explain, in part, why the research completed by many clinical faculty is indeed not scientific in design or analysis and therefore adds little to our knowledge base.

To address this problem, several departments of internal medicine developed training programs designed to encourage internal medicine residents to become clinical investigators and subsequently faculty [38, 39]. Goldman [40, 41] described the Harvard program he developed to train prospective faculty in clinical investigation. This program offers any physician the basic instruction in the fundamental principles required to become a clinical investigator devoted to patient-related clinical research.

The American Board of Internal Medicine now has a Clinical Investigator Pathway that allows a medical resident to enter a research training program after successfully completing 2 years of clinical resident training instead of the traditional 3 years. The research program must be 4 years long. The concept mandates that the clinical investigator spend 3 years in research training; 1 year can be spent doing the clinical work necessary to meet the clinical requirement of the subspecialty board in which the physician is enrolled. The research component need not be only basic laboratory investigation but could also include clinical research. The successful candidate, however, must have the specific training program approved by the American Board of Internal Medicine, must have a mentor, and must provide annual reports to assure that research training is progressing as planned.

Others have expressed concern about the lack of clinical investigation by medical school faculty and have made proposals for ways to improve the situation. What has been uniformly noted is the lack of training designed specifically for the physician wishing to enter clinical research [42-45]. The number of clinical investigators in ophthalmology has decreased at such a rate that a recent editorial by Epstein [46] lamented that the clinician-scientist in this specialty may no longer be a viable professional alternative for ophthalmologists. In 1979, Wyngaarden [47] called the clinical investigator an "endangered species".

Financial Future of Departments of Medicine

Insufficient clinical income is available to support both clinical faculty and the basic research faculty of a department. Departments of medicine should grow with caution and only after careful consideration of the financial consequences. A financial crisis will occur in many departments within the next few years, and it will be most severe in research-intensive departments that have expanded unchecked because of the availability of extramural funding. Once a few senior faculty who have survived on large grants lose their funding, there will be expectations that may be difficult to achieve. Two facts are worth remembering: No one is funded forever and faculty never retire when they are fully funded. The health care industry is changing rapidly, and every constituent has stated that the cost of health care is too high. When cost controls begin, faculty billings will not continue to increase at the unprecedented rates of the past decade. Most departments have already eliminated much of the inefficiency of documentation, billing, and collecting for clinical activity. Therefore, when reimbursements fall, the decision will not be what percentage increase in salary faculty will receive, but rather what percentage decrease they will tolerate. In this climate of confusion and contraction, the young physician-scientist can see little to encourage and sustain the drive to become a faculty member.

Desire of Medical Students

Surveys of graduating medical students have confirmed the impression that most new physicians do not want to begin a research career. Thirty-nine percent of senior medical students in 1960 said they were interested in a combined research, teaching, and practice career [48], but by 1982 as few as 22.2% expressed an interest in an academic career [34]. In 1990, 28.8% of graduating seniors, or 4443 new physicians, expressed a desire to pursue this type of career [12]. In 1963, 49% of Harvard medical school graduates gave research a high priority, but by 1976 only 2% considered research an important career choice [49]. A similar study from Iowa also found that only 2% of their medical graduates were interested in a career in research, teaching, or both [50, 51]. Of the 15 427 U.S. medical school graduates in 1990, only 1.3%, or 200, said they wanted a career that combined basic research and teaching [12]. This low percentage is not surprising because 70% of all medical school graduates' training in research is nonexistent or poor [52].

None of this is surprising. However, it is important to remember that a medical school's primary purpose and original founding intention was to train physicians to care for people; to remain on a school faculty was an acceptable but not the primary goal of the educational program.

Evaluation of internal medicine faculty confirms these impressions [13]. Of the current internal medicine research faculty, 59.5% received their training at a top 40 research-intensive medical school. In addition, of all internal medicine faculty who are doing research, 57.3% are members of these same 40 institutions. Therefore, most physicians are not exposed to research careers during their medical school training. The fact that 56.5% of faculty who consider themselves to be physician-scientists had research training during medical school supports the contention that research exposure during medical school affects a student's subsequent choice to become a clinical investigator or physician-scientist.

The Howard Hughes Medical Institute recognized the importance of these facts and developed two programs designed to assist the medical student who wants to become a physician-scientist. In collaboration with the NIH, they created the Research Scholars Program, which brings 35 medical students each year to the NIH for intensive training in an NIH laboratory. They also award individual research scholarships to medical students to pursue 1 year of research training at their own school. For those who are successful during that single year, further research training support is possible and even financial assistance for the rest of their medical education. This particular program offers not only the opportunity for research training during the formative period of medical school but also provides a method of debt reduction by continuing financial assistance during formal medical school training [53]. At its maturity, this program will support 300 students.

Future Faculty Opportunities

Is there room for new research faculty? In 1990, there were 615 412 physicians in the United States, and 16 930 were involved in research. The peak number doing research was 23 268 in 1985 Figure 6 [12]. Of the 131 172 internal medicine physicians, 7825 have a research career. In 1991, there were 12 892 faculty physicians in internal medicine departments, an increase of 38 faculty compared with 1990. In 1987, there were 11 293 full-time medicine faculty [13], a rate of increase of 312 per year from 1987 to 1991. Based on these trends, internal medicine faculty growth is slowing.

View larger version (22K): [in this window] [in a new window]   Figure 6. Physicians in research compared with the number of physicians from 1975 through 1990 in the United States [12] .

What are the faculty needs of internal medicine departments? If there is a 5% loss each year and it is desirable to have a 5% growth in faculty, then 1289 new faculty would be needed each year. Assume 50% will be clinical investigators (in either laboratory or patient-related research). Then there would be 645 research faculty positions in the United States in medicine departments. In 1990, 200 graduating seniors said they wanted to become basic investigators and 4443 wanted to become clinical investigators. Furthermore, last year the NIH sponsored 2798 physicians in research training, which is about 50% of all trainees. Therefore, last year there were 5596 physician trainees. If the training period is 3 years, then one third of this total, or 1865, would be available for a faculty position each year. Will we have enough room for these new faculty? What happens if our departments enter a no-growth phase? We are concerned about the future of the clinical investigator, both basic and patient-related, but we must develop a more defined plan for the future. This plan must address an improved method of training for students, house officers, fellows, and faculty. Faculty growth and attrition must be carefully discussed. Finally, we must develop a financial strategy that can support both the academic and the country's needs for health research. The existing funding of departments of medicine is in jeopardy and so is the future of the clinical investigator.

Recommendations for Improvement

1. To optimize national resources and maintain global leadership position in health care and biotechnology, a national medical research strategy must be developed that has both short- and long-range objectives. This requires identification of funding sources that will encourage stability and provide the opportunity for physicians to consider research as a viable professional goal.

2. As part of this strategy, consideration must be given to providing more funding to support both training and projects in clinical research if we are to translate laboratory discoveries into improved health.

3. The process of reviewing and awarding grant applications by the NIH, as well as other granting agencies, must be improved to reduce the syndrome of multiple grant application submission.

4. A rigorous research educational component that teaches the methods of both clinical and basic research must be developed in medical schools, residencies, and fellowship programs.

5. A mentor system that encourages the new physician-investigator must be given priority. Each school should create guidelines for mentors to assist them with their responsibilities. Being a successful mentor should be part of the faculty-retention evaluation process.

6. The separation of departments into clinical and research faculty should not be pursued. We must encourage sharing and collaboration among ourselves; otherwise, what is unique to medical research will be lost.

7. Departments must refrain from hiring new physicians to treat patients and allow the senior faculty more time for research. If this is not done, the force that will ultimately destroy the research ethos of our faculties will be empowered.

8. Chairpersons should spend time with their faculty, especially the junior faculty, fellows, residents, and students. If we wonder why interest in research is waning, part of the problem may be ourselves. We need to be visible; through discussion and careful listening, we can discover what needs to be changed. It is our duty to be in touch with clinical investigators, our students, and young faculty.

Challenges are to be conquered. We achieve excellence through problem solving. The problems we can clearly identify should not be ignored but rather skillfully overcome. There will always be a new challenge: As we learn more, we see more. My advice to new investigators is to keep your intellectual boundaries broad, seek a training setting that teaches the skills necessary to begin your quest, and never forget you are a physician. Regardless of the difficulty of the challenge, look from side to side periodically to view the other challenges that surround you because opportunities exist only where you look. And never be so goal oriented that all else in life is ignored.