Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 Nurturing Scientific Talent Maintaining a cadre of highly talented health scientists is the most critical element in sustaining the vitality of the U.S. system of health sciences research. Evidence from across the educational spectrum indicates that the United States is facing a future shortage of qualified researchers, which will threaten the nation's ability to prepare for scientific challenges of the twenty-first century. In the next 15 years many of the individuals who conceived the ideas that have revolutionized health sciences research will be retiring. Neglect in educating and training their replacements inevitably will lead to a decline in the nation's capabilities in health-related research, an area in which the United States has maintained an unchallenged world leadership for the past 40 years. Particularly alarming is the apparent decline in the number of physi- cians engaged in health-related research. The study of many fundamental biological questions begins with investigation into human disease processes, and human data are essential to address these questions effectively. The defining and understanding of these problems are largely in the hands of physician-scientists, who also serve as technology-transfer agents, translating fundamental laboratory discoveries into clinical practice. Accurately assessing the magnitude and timing of an impending per- sonnel shortage depends upon a variety of factors. Scientific employment growth in academia, government, and the private sector is tied closely to the economic health of the nation. As the post-World War II baby boomers grow older, the retirement rate among scientists trained in the l950s and 1960s will accelerate, increasing the demand for replacements. Also, a 117
118 FUND NO HEALTH SCIENCES RESEARCH higher death rate in this more elderly scientist population will increase demand in the health scientist labor market. The composition of the future health scientist work force also will be affected by changing demographics with regard to age, gender, race, ethnicity, and immigration, as well as the quality of scientific education and training. An accurate assessment of all of these factors affecting the scientific work force must be part of a decision-making process regarding research training needs. A comprehensive talent renewal plan must encompass the multitude of research disciplines that range from basic to applied investigation. This chapter examines the available data on the research work force and highlights the possible implications for the future health scientist talent pool. PROBLEMS IN THE HUMAN RESOURCE BASE Precollege The pathway to a scientific career does not begin in undergraduate or postgraduate years; rather, an interest in science is kindled in the early years of formal education kindergarten through grade 12. However, several recent national and international studies have shown a continuous decline in science and mathematics skills by American students at all educational levels. Although the committee focused its deliberations on the resources for graduate and postdoctoral education and training, the committee recognized that competency in precollege and undergraduate science and mathematics education is critical for preparing students for scientific careers. Additionally, an early appreciation of the excitement of scientific discovery is important for attracting students into scientific careers. About three-quarters of all students who eventually major in science and engineering follow a college preparatory curriculum.) However, the National Assessment of Educational Progress, which is part of the federally sponsored Nation's Report Card conducted by the Educational Testing Ser- vice, concluded that only 7 percent of 17 year olds in 1986 were prepared adequately for college-level science courses.2 This report also confirmed the race and ethnicity gaps that previous studies have found in science achieve- ment. Whereas only about 15 percent of African-American and Hispanic 17 year olds demonstrated the ability to analyze scientific procedures and data, nearly half of their white peers could do so. ~ . Despite inherent difficulties in interpreting comparative international education data, a study of mathematics and science abilities among students in four foreign countries, four Canadian provinces, and the United States ranked the American students near the bottom in these skills.3 The National
NURTURING SCIENTIFIC TALENT NUMBER (Millions) 3 2.9 2.8 2.7 2.6 2.5 / \ Rae / \ 2.4 , , , 1 , , 1 , 1 , 1 1 1 1 1 , 86 87 88 89 90 91 92 93 94 95 96 97 98 99 0 1 2 YEAR 119 r ... . ... . . 3 4 FIGURE 5-1 Projected number of U.S. high school graduates from 1986 to 2004. (Repented with permission from Western Interstate Commission for Higher Education. High School Graduates: Projections by State, 1986 to 2004. Boulder, CO; 1988.) Research Council has drawn attention to the poor mathematics proficiency of American students in a report entitled Everybody Counts: A Report to the Nation on the Future of Mathematics Education.4 This study called for a complete overhaul of precollege mathematics education in the United States and suggested alternative educational strategies to counteract this growing problem. Moreover, national demographic evidence indicates that the number of high school graduates is expected to decline by 12 percent between 1988 and 1992, from nearly 2.77 million to 2.44 million students (Figure 5-1~.5 Unless these trends change, the declining number of high school graduates is expected to lead to declining undergraduate enrollment in U.S. colleges and universities in the early to mid 1990s. However, the number of high school graduates is expected to return to the 1988 level by 1998 and will coincide directly with a rapidly increasing retirement rate of university faculty. Undergraduate A recent study by the Office of Technology Assessment, tracldng the progress of American students toward careers in science and engineering,
120 FUNDING HEALTH SCIENCES RESEARCH exemplifies the attrition rates from the available talent pool during all stages of scientific career development.6 From an original study group of 4,000 ninth-grade students, only 1,000 had sufficient mathematics abilities at that point to pursue a scientific or engineering career. When these students had completed their secondary school education, only 500 were adequately prepared to continue in a science or engineering college curriculum. At this point, women were represented equally in the study group. However, upon entering college the number of women electing to pursue a science or engineering career fell to 44 of 250 individuals compared to 140 of 250 for men. By the completion of their baccalaureate programs, only 66 of the original study group of 4,000 received B.S. degrees-a precipitous drop of more than 98 percent in the original. This example illustrates vividly the problem of recruiting individuals into the sciences, especially as it applies to women and other underrepresented groups. Unfortunately, major losses in the science and engineering talent pool occur during the undergraduate years.7 Students usually make career deci- sions during this critical undergraduate period. Thus, recruiting individuals into the health sciences will depend upon the following factors: · enthusiasm engendered by high-quality teaching, · scientific opportunity and excitement, · economic status of the nation, financial support for education, and financial rewards from employment opportunities. Students interested in health sciences research generally follow curric- ula that prepare them for graduate study leading to professional degrees either Ph.D.s or M.D.s. Relevant areas include not only biology and chem- istry but also such fields as physics, mathematics, psychology, or the social sciences. Data gathered over the past 10 years reveal a decline in earned bachelor degrees in the life sciences (Figure 5-2~.7~8 Although life scientists are not the exclusive talent pool for the health sciences, the committee believes that a significant portion of health scientists with advanced degrees come from this student population. Also, the committee believes that over the last 10 to 15 years, the supply of high-quality graduate students in the health sciences has declined. Considerable discussion has focused on reasons for the failure of science educators to stimulate student interest in scientific careers. The committee believes that this failure may be due partly to the widespread practice of collegiate science education stressing passive learning through lectures rather than active learning through participation in research. Hon- ors programs that include hands-on research in conjunction with faculty mentors provide an example of active learning that can stimulate students
NURTURING SCIENTIFIC TALENT 250 200 150 100 50 121 Thousands if+ O 1 1 1 71 73 75 77 79 81 83 85 YEAR Life Sciences ~ Soc/Behav Sciences - + - Busine&s/Mgm t FIGURE S-2 Number of bachelors degrees awarded in the life sciences, social and behavioral sciences, and business/management from 1965 to 1985.3 to pursue research careers. Also, family values concerning education have a great deal of influence on childhood learning and performance. Although science education and training for undergraduates fall within the purview of the Science and Engineering Education Directorate of the National Science Foundation (NSE;), the problem of recruiting students into science and engineering careers has recently been addressed by Congress. The National Science Scholars Program, part of the President's Educational Excellence Act, is designed to encourage exceptional students to pursue careers in scientific and engineering fields.9 Modeled on congressional appointments to military academies, the proposed program calls for federal support for undergraduate education in science and engineering for two appointees (one female and one male) for every member of Congress. The awards would be 4-year fellowships, based on merit and a competitive selection process, for study at an institution of the student's choice. The program would sponsor approximately 1,000 new scholarships per year, each having an annual stipend of $5,000. Staling $18 million per year when fully operational, this program should act as a catalyst to attract additional student financial aid from other sources. The committee believes that this type of program focuses local attention on science and engineering education and serves as a highly visible example of congressional support for
122 FUNDING HEALTH SCIENCES RESEARCH renewing scientific talent. The small size of this program, however, clearly will not be sufficient to meet expected shortages in all of the sciences. The racial and ethnic composition of the U.S. population is changing, and these changes also will affect the pool from which scientific talent is drawn. The U.S. Bureau of the Census estimates that the minority composition of the 22-year-old population will grow to 20 percent by 200W up from 14 percent in 1975.~° Ethnic and racial minorities historically have been underrepresented in science and engineering. Whereas nearly 5 percent of white and Asian 22 year olds have earned baccalaureate degrees in natural science or engineering, only 1.6 percent of blacks, Hispanics, and Native Americans have earned the same degrees. In the 1970s the National Institutes of Health (NIH) created the Minority Access to Research Careers (MARC) Honors Program to increase the number of minority students pursuing graduate study leading to a Ph.D. in biomedical science. The largest portion of the MARC program is the Honors Undergraduate Research Raining Program. Trainees at selected institutions receive tuition support and a stipend to participate in a specially structured curriculum. Working closely with faculty members on laboratory research projects in the biomedical sciences is a key element in the training experience. Longitudinal data are not yet substantial enough to determine if this program is having a significant effect on recruiting minorities into graduate health research programs, however. Although the Institute of Medicine (IOM) undertook a survey evaluation of the MARC program in 1985, it was too early to gauge the success of the program, and no remedies were suggested to address the problems identified. Scientific Doctorates (Ph.D.s) The transition from undergraduate to graduate school is another criti- cal juncture in the retention of candidates for future careers in the health sciences. Large numbers of undergraduates elect not to pursue graduate studies; of those who do, an unknown number either may not complete their graduate program or may leave research upon earning a doctorate. Data from the NSF Survey of Graduate Science and Engineering Students and Postdoctorates reveal that graduate student enrollment (not including postdoctorates) in the life sciences has grown only slightly since 1980, from 102,504 to 108,641, whereas enrollment in the physical sciences has grown by more than 23 percent annually, from 26,952 to 33,203.~3 Enrollment in the social sciences has declined slightly over the same period, from 94,778 to 91,884.~3 There is a considerable lag time affecting the scientific labor market that must be considered when policy is formulated. Presently recognized opportunities will affect the scientific work force 5 to 8 years hence, upon
NURTURING SCIENTIFIC TALENT 123 completion of a doctoral program, clinical training program, or a postdoc- toral fellowship. In short, salaries and economic opportunity in 1989 will have affected graduate enrollment that year, but the 1989 graduate school entrants will not affect labor force supply until possibly 1995. This scenario is compounded by the fact that there also will be a significant decline in the size of the 18- to 24-year-old population the talent pool available for recruiting into graduate study in the late l990s, when many tenured faculty members are expected to retire.8 The number of foreign graduate students enrolled in U.S. institutions as well as the percent of degrees conferred on foreign nationals has in- creased steadily over the past decade.~4~5 The United States produces nearly 14,500 natural scientists and engineers annually, up from 12,000 in 1978. In 1987 about 9,700 science and engineering doctorates were conferred on U.S. citizens or foreign nationals with permanent visas. The remaining 3,800 doctorates were conferred on foreign citizens with tempo- rary U.S. visas. Of those students on temporary visas receiving doctorates in 1987, about half remained in the United States to pursue employment (23 percent) or postdoctoral studies (25 percent). Thus, immigration com- pensates for shortages in trained U.S. personnel and adds to the intellectual and technological abilities of the country's scientific work force. But while foreign students earned only 16 percent of the doctorates in life sciences in 1987, the committee believes that the health sciences should not follow the path of engineering, in which almost half of the doctorates are conferred on foreign nationals. Such heavy reliance on foreign talent could jeopardize the future success of American science efforts and the national economy should fewer and fewer degree recipients elect to remain in the United States. Although women make up 42 percent of the U.S. work force (U.S. Department of Labor Statistics, personal communication, 1~19-89) they have been underrepresented historically in science and engineering. In 1977 women represented only 10.4 percent of all doctoral scientists and engineers.~7 Although their numbers have grown from 31,800 in 1977 to 73,423 in 1987, for example-women scientists and engineers still account for only 16.3 percent of the total doctoral population. However, the annual proportion of doctorates conferred on women has been growing steadily over the last three decades.~5 In 1987 women earned 35 percent of the doctorates awarded in the life sciences. In the social sciences they earned 43 percent, but they earned only 17 percent of all doctorates in the physical sciences. Since individuals in each of these broad categories pursue careers in the health sciences, these data indicate that there have been small gains toward equal representation between men and women in the scientific work force. The proportion of non-Asian minorities receiving doctorates is not
124 FUNDING HEALTH SCIENCES RESEARCH increasing; rather, recruitment appears to be worsening in these groups.l5 Whereas the number of black women earning doctorates annually between 1977 and 1987 has remained fairly steady at about 500, the number of degrees conferred annually on black men has been halved to about 300. Also, whereas doctorates conferred on Hispanic women have more than doubled to 28~in the same period, the number going to Hispanic men has remained steady at about 300. These data raise serious questions about policies and programs for improving minority participation in higher education as well as research and pose a problem regarding cultural values. The committee emphasizes that the recruiting difficulties of non-Asian minority males should be of particular concern to all policymakers and educators. The United States employs about 12,500 new Ph.D. scientists and engineers each year. Industry has been creating about 5,500 new Ph.D. positions per year for scientists and engineers. If these hiring practices prevail and retirements in this sector begin to increase as we approach the year 2000, the demand in industry and business could increase to nearly 9,500 by that time. Retirement rates of academic faculty also are expected to increase over the next 15 years, rising from about 2,000 in 1988 to more than 4,500 in 2004. Although demographic evidence indicates that there may be a dearth in undergraduate enrollment in the early l990s, the impending retirements, coinciding with a surge of 18 to 24 year olds toward the end of the next decade could create an annual academic demand for new Ph.D. scientists and engineers of nearly 8,500 by 2004. At current production rates, even if we rely heavily on the possibility of filling positions with foreign students receiving U.S. doctorates, the annual shortfall still may be as high as 7,500 in the first decade of the twenty-first century. Although these data predict shortages for all natural sciences and for engineering, potential shortages in the health sciences can be expected as well. Data from the Doctorate Records Survey shows that between 1973 and 1987, employment of biomedical scientists by all sectors grew 4.9 percent annually, rising from 43,000 to 84,500. This includes the 43,000 scientists and 8,200 postdoctorates employed by academic institutions regardless of their level of research activity, 16,000 scientists employed by industry, as well as other Ph.D.s outside academia actively engaged in or managing research and development. Seventy-six percent of these scientists hold doctorates in biomedical sciences (Figure 5-3~; the remaining twenty-four percent have doctorates in fields other than biomedical science. Over this same period the annual output of new biomedical Ph.D. recipients grew by 12.8 percent from 3,520 to 3,969. However, not all recipients of biomedical degrees are employed as biomedical scientists; approximately 24 percent are engaged in other activities (Figure 5-4~. Over the past decade the growth in employment for
NURTURING SCIENTIFIC TALENT / /:: f Biomedical Sciences 7696 :: 125 Other 996 Physical Sciences 8% ,7 Engineering 3% \ I ,~ Behavioral Sciences 4% Degree Types as a Percent of the Workforce FIGURE 5-3 Composition of the biomedical work force. (Source: Office of Science and Engineenng Personnel, National Research Council) Biomed ical 76% a_ Other 24% ~ _ Field of Employment Physical 3 Engineering 1.2 Humanities 3.5 Behavioral 0.5 Other 14.6 FIGURE Sot Employment of biomedical scientists. (Source: Office of Science and Engineering Personnel, National Research Council) biomedical scientists largely has been in industry, growing an average of 12 to 13 percent annually.l9 The employment of behavioral scientists grew 113.6 percent between 1973 and 1987, rising from 31,669 to 67,651.~9 More than 91 percent of the vacancies in behavioral sciences are filled by individuals with doctorates in the behavioral sciences. The number of behavioral sciences doctoral degrees conferred annually has climbed from 3,542 to 3,960, reflecting an 11.8 percent change over the same period. One element that may skew these data is the surge in clinical psychology degrees that has occurred over the past few years. One factor affecting the output of Ph.D. scientists is the increasing time needed to earn a doctorate.20 Whereas the median registered time to degree (i.e., the time the student is registered for formal courses or thesis preparation with the university registrar) for all fields was about 5.4 years in 1967, by 1987 it had increased to 6.9 years. While the largest increase was noted for graduate students in the humanities, doctoral candidates in the physical and life sciences now spend more than 6 years in graduate study, compared to just over 5 years two decades ago. In the social sciences the median was 7.2 years in the 1987 survey, compared to 5.2 years in
126 FUNDING HEALTH SCIENCES RESEARCH 1967. It is not clear to the committee how this affects the financial support mechanisms from the federal government or other sources. Professional Doctorates (M.D.s and M.D./Ph.D.s) Physician-scientists are charged with carrying fundamental discoveries in the laboratory to the patient and assessing the efficacy of new treatments and other interventions for improved health care. The recruitment of physician-scientists into research careers is hampered severely by the length of time necessary for clinical training, the often unfocused structure of clinical research experience, the need for the individual to understand increasingly complex technologies, and the requirement of the physician to generate clinical income at the expense of time for performing research. At a time when biology and medicine offer exciting opportunities for improved health care, this declining interest in investigative careers is particularly troublesome. The problem is magnified for the fields of public health and preventive medicine, where no practice income is raised to support salaries or subsidize education. The majority of M.D. and M.D./Ph.D. scientists are employed by med- ical schools, the government, and private research institutions. According to the American Medical Association (AMA) Physician Masterfile, there were 569,160 federal and nonfederal physicians in the United States as of December 1986.2i Of these, 86,670 (15.2 percent) were female and 123,090 (21.6 percent) were foreign medical graduates (excluding Canadian gradu- ates). The number of physicians reporting research activity had grown from 11,929 in 1970 to 18,535 in 1983. However, from the time of the 1983 survey to 1986, there was a drop of nearly 700 to 17,847 physicians engaged in research.2i This also reflects a drop from 3.6 percent of the total physician population engaged in research in 1983 to 3.1 percent in 1986. The 1986 population of physician researchers was composed of 16.2 percent women and 23.5 percent foreign medical graduates, both groups being represented slightly higher than their proportion in the total physician population. A1- though these data may be flawed and the small shifts reported by the AMA may not be significant, the committee believes that in recent years there has been no growth in the number of physicians participating in research. The number of applicants to U.S. medical schools has dropped by more than 30 percent in the past 10 years, from 40,600 in 1977 to 28,100 in 1987 (Figure 5_5~.22,23 This decline has engendered concern in the nation's medical centers about the future quality of medical care in the United States as well as the capabilities of physician-scientists. With the increasing sophistication of health sciences research, educa- tors have recognized the need to develop pathways to ensure that physicians
NURTURING SCIENTIFIC TALENT NUMBER (Thousands) 40 30 20 10 127 77 78 79 80 81 Total Applicants 82 83 84 85 86 87 YEAR _ Total Enrollees FIGURE 5-5 Number of U.S. medical school applicants and enrollment Mom 1977 to 1988 22,23 are as rigorously trained in scientific methodology as their Ph.D. counter- parts. One pathway for achieving this goal is to encourage some physicians to enter doctoral programs in specific research areas leading to a combined M.D./Ph.D. degree. While more than 100 of the 127 U.S. medical schools offer programs for combined M.D./Ph.D. degrees in various areas such as biomedical sciences, social sciences, humanities, biomedical engineering, and law, and only 20 to 30 graduate significant numbers of M.D./Ph.D. candidates.23 Such combined training provides enhanced research experi- ence that more thoroughly prepares physician researchers for independent basic or clinical investigation. Some committee members believe that although M.D./Ph.D. programs provide a suitable model for training physicians in research methodology, these are not the pathways followed by most physicians pursuing careers as independent investigators. There are existing models in the nonbiological sciences that tailor coursework in areas to meet the special needs of the physician-scientist, and that link supervision with an established physician mentor (e.g., the Robert Wood Johnson Clinical Scholars Program). For physicians who choose investigative careers in disciplines such as epidemi- ology, health services research, or health policy, these alternative models may be preferred.
128 FUNDING HEAI=H SCIENCES RESEARCH From Degree to Scientist The prolonged period of time it takes to earn a doctorate and the subsequent extensive postdoctorate training time necessary for both Ph.D. and M.D. scientists often force these individuals to postpone at least some aspects of their personal lives. Both the financial concerns of young families and the balancing needs of two~areer families encourage these scientists to move more quicldy to establish a stable career. Clinical training is of particular concern because it requires a sub- stantial time investment, especially if the physician embarks on a career requiring subspecialization. Subspecialties in internal medicine and surgery now require between 5 to 7 years of postdoctoral training after medical school. Since many specialty boards do not allow credit toward certification for research, a formal research training period most often following the clinical subspe~cialty training, extends the training time invested to 8 or more years. This training time often coincides with the payback period of the considerable financial debt that many physician graduates accumulate during medical school. Moreover, because of clinical training demands, research training ex- periences for physician-scientists often are unstructured and poorly focused. It is rare for either clinical training or clinical research experiences to in- clude formal instruction in scientific design, research methodology, and statistical analysis. Additionally, if they lack critical review or accreditation, clinical research training programs fail to introduce standards and account- ability. As a result, physician-scientists often are less prepared for pursuing research than more rigorously trained Ph.D. scientists who have had 4 to 5 years of formal research laboratory training. A 2-year research experi- ence, particularly when poorly focused, often leaves physician-scientists less prepared for competing in the peer-reviewed grant system than are more formally trained Ph.D. scientists. Other pressures in the modern medicine environment add to the discouragement of physicians involved in clinical investigation as well. A recent IOM report on resources for clinical investigation concluded that fundamental changes in the organization of health care and the mounting efforts aimed at cost containment discourage clinical research scientists from pursuing clinical investigations.24 Along with the pressures that young physician-scientists face earn in their careers, there are pressures upon all physicians to earn clinical income for their academic health center. Clinical income is more predictable than research grants, particularly in terms of institutional revenues. As medical schools rely more and more on faculty practice plans for salary support, clinical faculty members are pressured to maintain their clinical practice incomes. These pressures~irect or indirect, bold or subtle are felt by virtually all M.D. investigators.
NURTURING SCIENTIFIC TALENT 129 Compounding these difficulties, the practice of medicine has become more complex and uses more advanced technology than ever before. Even the so-called cognitive specialties such as internal medicine are heavily de- pendent upon advanced technological procedures, which require technical skills that must be practiced regularly to maintain a high level of compe- tence, making it more difficult for physician-scientists to devote precious time to scientific investigation (unless these individuals are in unusually supportive academic environments). PROBLEMS WITH THE FINANCIAL SUPPORT BASE For nearly 40 years the Science and Engineering Education (SEE) Di- rectorate of NSF has been the primary sponsor of programs for developing scientific talent at the undergraduate level. At its peak in 1960 and 1961, this directorate controlled more than 40 percent of the NSF budget.25 In the ensuing 20 years, appropriations to SEE failed to keep pace with other parts of the NSF budget, until only 1.5 percent was allocated to science and engineering education by 1983. In recent years, however, the admin- istration has recognized the vital importance of science and engineering to national security and international competitiveness. This reemphasis is reflected in the recent NSF budgets where funding to SEE has grown from $55 million in 1987 to a proposed $251 million in 1991 nearly 10 percent of the 1991 NSF budget. Federal support for training health scientists began with the passage of the National Cancer Act of 1937 which authorized the U.S. Surgeon General to provide fellowships and train personnel for cancer research and prevention. This authority was expanded in the Public Health Service Act of 1944, expanding training programs sponsored by the NIH. This act not only increased the research capacity of the U.S., but also provided broad financial support to medical students, whether or not they expected to pursue research careers. In 1973 the Nixon administration impounded NIH training funds in an effort to phase out all research training. Congress responded by passing the National Research Service Award (NRSA) Act (P.L. 93-348) in 1974. This act authorized training at the level of the Public Health Service to be conducted primarily in the NIH, the Alcohol, Drug Abuse, and Mental Health Administration (ADAMHA), and the Health Resources Service Administration (HRSA). By creating a separate authorization, research training is now loosely connected to research but the budgets are acted upon separately by congressional appropriations committees. The NRSA act eliminated support for medical students except those pursuing research careers. Additionally, the act included a service obli- gation requiring those trainees receiving funds to be actively engaged in
130 250 200 150 100 50 FUNDING HEALTH SCIENCES RESEARCH DOLLARS (Millions) 350 300 _ ~ , O 1 1 1 1 1 1 1 1 1 1 1 1 1 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 YEAR - Constant 1988 $ -+- Current $ {~.) {~es.) FIGURE 5~ NIH obligations for National Research Service Award (NRSA) training. (AppendLx liable A-12) research equivalent to one month of service for each month of support. This requirement has been modified to allow for short periods of support without a payback However, if trainees elect not to pursue research ca- reers they must pay back the costs of their education to the government. The NRSA act limited support to an aggregate of 5 years for predoctoral studies and 3 years for postdoctoral research. The NIH and ADAMHA are the primary federal sponsors for training in the health sciences. In 1971 NIH allocations for training as a percent of R&D funds exceeded 18 percent.26 Research training allocations fell below 11 percent of the NIH research budget in 1973 and have continued to decline, accounting for less than 5 percent of R&D allocations in 1988. Additionally, appropriations targeted for training have declined from nearly $290 million in 1980 to about $250 million in 1990 when measured in constant 1988 dollars (Figure 5~.27 The number of full-time training positions (t l l Ps) supported by NIH has remained fairly constant each year-between 11,000 and 12,000 since the late 1970s (Figure 5-7~. However, in order to increase sagging stipend levels, NIH trimmed support for 1,000 ~ l lPs in fiscal year 1989.28 Since NIH supports approximately one-quarter of the graduate students in the biomedical sciences through the NRSA program, these cuts in training
NURTURING SCIENTIFIC TALENT 131 positions were quite significant. NIH reestablished these positions by re- programming other funds in 1989. Declining training support has devastated the training of the next gen- eration of behavioral and social scientists.29 In the early 1970s ADAMHA allocations for research training exceeded 14 percent of R&D funds. As with NIH, training allocations in ADAMHA have declined to about 5 per- cent of research funds in 1988. Whereas NIH training obligations have declined about 17 percent in constant dollars, ADAMHA obligations for research training have been reduced by more than half since 1977 (Figure 5-8~. There also has been a concomitant decline in the number of training positions, falling from 1,800 in 1977 to a low of 1,100 in 1986. The number of positions rebounded slightly, to nearly 1,300, in 1988 (Figure 5-9~. Raining funds from NIH and ADAMHA are awarded through compet- itively reviewed institutional training grants or individual fellowship awards. About 85 percent of NIH-sponsored training appointments are supported on NRSA training grants awarded to institutions for either predoctoral (50 percent) or postdoctoral (35 percent) training.27 Of the remaining training funds, 13 to 14 percent are awarded through NRSA individual postdoctoral fellowship awards, and slightly less than 2 percent are allocated to individ- ual predoctoral fellowships. About 55 percent of the predoctoral training positions are awarded through the NIGMS followed distantly by NCI with 14 1 2 10 8 6 4 2 o NUMBER of FTTP (Thousands) 77 78 79 80 81 82 83 84 85 YEAR ~ Predoctoral ~ Postdoctoral 86 87 88 89 90 91 (...., ,..t., FIGURE 5-7 Number of full-time training positions (t l lYs) sponsored by the NIH from 1977 to 1991. (Appendix Able A-19)
132 35 30 25 20 15 10 FUNDING HEALTH SCIENCES RESEARCH DOLLARS (Millions) o 78 79 80 81 82 83 84 YEAR ~ Constant 1988 $ ~+- Current $ 85 86 87 88 89 FIGURE 5~ ADAMHA obligations for NRSA training. (Appendix liable A-20) 2000 1 500 1 000 500 o NUMBER of FTTPs _-, i,:.:.: ... . .. 78 79 80 81 82 83 84 YEAR Predoctoral 85 86 87 88 89 Postdoctoral FIGURE 5-9 Number of full-time training positions (LllPs) sponsored by ADAMHA from 1978 to 1989. (Appendix Table A-20)
NURTURING SCIENTIFIC TALENT 133 10 percent of the predoctoral slots. NHLBI has the largest portion of postdoctoral positions-sponsoring about 20 percent of all NIH-supported postdoctorates. Similar to NIH, about 87 percent of ADAMHA training funds are distributed through predoctoral and postdoctoral training grants, and only 13 percent support fellowship awards.30 Currently, there is about equal dis- tribution between predoctoral and postdoctoral support of full-time equiv- alent training positions in both NIH and ADAMHA Whereas this ratio has been stable over the last decade for NIH, the cuts to research training in ADAMHA have affected only predoctoral positions, which have fallen from 1,178 to 694.29 30 It should be noted that ~ l l Ps totals are generally less than appointments because several short-term appointees can equal one ~ l lP. The distribution between M.D. and Ph.D. postdoctoral training ap- pointments has shifted slightly since 1980. In 1980 support was weighted more heavily toward Ph.D. postdoctorates, with 3,656 supported in com- parison to 2,092 M.D. postdoctorates.27 By 1987 the number of M.D. postdoctorates had increased to 2,532, thereby bringing support more in line with the 3,139 Ph.D. postdoctorates supported that year (Figure 5-10~. Increased efforts to support more physician-scientists should increase the competitiveness of this group and enable them to win a larger share of investigator-initiated research project support. The Medical Scientist Gaining Program (MSTP) sponsored by NIH is the largest national program for individuals pursuing joint M.D./Ph.D. degrees. This program is sponsored by NIGMS and has supported about 700 MSTP trainees annually throughout the 1980s.27 Although the NIH funds programs in 28 medical schools, many more combined programs are supported in U.S. medical schools by private, state, and institutional funds. However, the committee was not able to determine the size of these commitments.23 The NSF Survey of Graduate Science and Engineering Students and Postdoctorates reports that large numbers of students are supported by teaching assistantships and a smaller but still significant number are sup- ported on research grants.~3 Unfortunately, graduate students and postdoc- toral fellows supported on research project grants from NIH and ADAMHA are not identified in the NIH database and the magnitude of this support, therefore, is difficult to ascertain. However, data from the Survey of Grad- uate Science and Engineering Students and Postdoctorates conducted by the NSF indicates a growing trend toward supporting trainees as research assistants on NIH research grants. The recent NRC report, Biomedical and Behavioral Research Scieniists: Their Raining and Supply, estimates that NIH supported research assistantships have grown from 2,673 in 1979 to 4,426 in 1987.
134 14 12 10 8 6 4 FUNDING HEALTH SCIENCES RESEARCH NUMBER OF APPOINTMENTS (Thousands) o 80 81 82 83 84 85 86 87 88 YEAR ~ M.D. ~3 Ph.D. ~ Predoctoral ~ Total FIGURE 5-10 Number of trainee appointments in NIH sponsored training programs by academic level from 1980 to 1988. (Appendix Bible A-21) The MARC program administered by NIH attempts to address the problem of underparticipation by minority groups in the health sciences at both the undergraduate and graduate levels of training. Since 1982, NIH has supported about 400 MARC undergraduate training positions annually.27 However, NIH support for MARC NRSA faculty fellowships has been dismal. In 1980 NIH supported only 36 of these faculty fellowships, and the number declined steadily to 18 in 1987. The committee believes that although this program offers the potential for recruiting individuals in minority groups into the health sciences, limited data do not allow a thorough program evaluation. Other Support Mechanisms By all measures, the private sector is increasing its commitment to training health scientists as well. According to one estimate, more than $17 million were invested in training by private foundations and voluntary health agencies. Contributions made at the undergraduate level generally provide support for curriculum development and improving the undergrad- uate teaching environment. For example, the Howard Hughes Medical Institute (HHMI) has initiated a series of grants programs to strengthen undergraduate science education and research in private undergraduate
NURTURING SCIENTIFIC TALENT 135 colleges and research universities with undergraduate colleges.31 The goal of this program is to increase the number of students, especially minori- ties and women, pursuing careers in the biomedical sciences. In 1988 HHMI awarded $30.4 million to 44 colleges, including 10 historically black colleges. In the second year of the program the Institute expanded this initiative with $61 million awarded to colleges affiliated with research uni- versities and other doctorate-granting institutions. Voluntary health agen- cies generally sponsor research fellowships or career development awards for postdoctoral training in their respective area of interest (e.g., cancer, heart disease, arthritis). Other programs target specific groups like the Robert Wood Johnson program to encourage underrepresented minorities to pursue careers in the health sciences including biomedical research.32 The committee believes that an increasing number of postdoctoral fel- lows are being supported by industrial sponsors. Favorable tax policy that has stimulated growing levels of investment in research and development may be responsible for this growing trend. Postdoctorates may be spon- sored directly by the pharmaceutical or biotechnology industries to work in industrial R&D laboratories or, in some instances, in academic settings. It is unlikely that industry will invest significant amounts of funds at the undergraduate or predoctoral levels of training without more assurances that these trainees will be employed by their firms. Like foundations, corporate contributions for undergraduate and predoctoral education and training most likely will be used for curriculum development and updating the teaching environment. However, no centralized data base is available to determine either the magnitude of industry and private nonprofit support or the number of individuals supported. Clearly, the private sector can play a very significant role in training future health scientists, but the committee believes it simply cannot replace federal funding for research training. SUMMARY AND CONCLUSIONS The committee emphasizes that the single most critical long-term investment in the U.S. health sciences research enterprise is the sustained development of well-trained, creative scientists. Future progress toward improving health will continue only if efforts are sustained by talented individuals on all fronts to ensure a balanced attack on disease processes and exploration of all means of disease prevention. The emergence of an unexpected health crisis such as AIDS emphasizes the importance of trained scientific personnel who can be redirected quickly as needed. Successful handling of future epidemics will require a strong health sciences research system, particularly trained researchers. Demographic data indicate that later this decade there will be in- creasing attrition of scientists trained in the 1950s and 1960s. Removing
136 FUNDING HEALTH SCIENCES RESEARCH mandatory retirement ages may reduce some attrition due to retirements, but the effects will not be measurable until many years later. Employment growth in the private sector over the past decade has exceeded that in academia twofold. If this trend continues, competition for scientific talent between academia and industry will intensify. Thus, evidence is mounting that the supply of health sciences re- searchers will be grossly inadequate to meet estimated demands by the end of this decade. These work-force trends will slow advances in the health sciences if they are not offset with careful planning and allocation of resources. In order to develop a highly qualified population of health scientists for the twenty-first century, the committee believes that: . at a minimum, steps must be taken now to maintain the pool of scientific intellect in our society by improving the quality of science education and training; · efforts must focus on recruiting, training, and retaining the most promising and talented individuals; and · any new strategies should include programs targeted at increasing the numbers of scientists from underrepresented groups as well as improving multidisciplinary and interdisciplinary training of scientists. Coordinated efforts across the educational spectrum are needed to sustain a pool of qualified health science researchers and to continue the progress already made in improving both the health care and quality of life of the American people. The failure to recruit qualified candidates into the health sciences is due partly to declining levels of support in the NRSA predoctoral and postdoctoral training programs as well as to neglect across the entire educational spectrum. The committee also concluded that the number of trainees supported on research project grants has been growing. Indeed, this type of support closely links research training with research. However, the committee acknowledges that there are disadvantages to supporting training on research project grants as well. Research grants commonly do not provide tuition support for graduate students since they may be classified as technical assistants receiving salary. The committee be- lieves that often times support for these positions are reduced or removed when study sections provide recommended funding levels. Also, supporting trainees on research grants obligates trainees to perform established re- search protocols in order to ensure research productivity for the principal investigators rather than acquiring a broad philosophical background for asking pertinent scientific questions. Policies therefore must be developed to address the needs of ongoing research as well as those ensuring the long-term vitality of the health sciences enterprise. The committee is convinced that allocation policies in recent years em- phasizing research project support have underemphasized the commitment
NURTURING SCIENTIFIC TALENT 137 for broad training experiences. Resource allocation policies should foster the development of highly qualified health researchers and should pro- vide the opportunity for support throughout their careers. These policies should focus on the long-term goals of the research enterprise rather than short-term corrections. Academia, government, and industry must play co- operative roles in developing and pursuing effective strategies for enhancing and renewing the nation's health sciences talent base. Furthermore, alloca- tion policies for training must prepare the nation for achieving its long-term research goals rather than merely making short-range adjusunents to meet current needs. REFERENCES 1. U.S. Congress; Office of Technology Assessment. 1988. Educating Scientists and Engineem: Grade School to Grad School. OTA-SET-377. Washington, D.C.: U.S. Government Printing Office. 2. Educational Testing Service. 1988. National Assessment of Educational Progress: The Science Report Card: Elements of Risk and Recovery. Princeton, N.J. Lapoint, A.E., N.A. Mead, and G.W. Phillips. 1989. A World of Differences: An International Assessment of Mathematics and Science. Princeton, NJ.: Educational Testing Service. 4. National Research Council. 1989. Everybody Counts: A Report to the Nation on the Future of Mathematics Education. Washington, D.C.: National Academy Press. Western Interstate Commission for Higher Education. 1988. High School Graduates: Projections by State, 1986 to 2004. Boulder, CO. 6. U.S. Congress; Office of Technology Assessment. 1985. Demographic [lends and the Scientific and Engineering Workforce. OTA-TM-SET-35. Washington, D.C.: U.S. Government Printing Office. 7. U.S. Department of Education, Office of Educational Research and Improvement. 1989. The Condition of Education, 1989: Postseconda~y Education. Volume 2. CS 89-651. Washington, D.C.: National Center for Education Statistics. 8. National Academy of Sciences: The Government-University-Industry Research Round- table. 1987. Nurturing Science and Engineering Talent: A Discussion Paper. Washing- ton, D.C.: National Academy Press. 9. Abelson, PH. 1989. Congressional fellowships for science. Science 243~4899~:1649. 10. U.S. Department of Commerce: Bureau of the Census. Projections of the Population of the United States by Age, Sex, and Race, 1983 to 2080. Current Population Reports Series P-25, No. 925. Washington, D.C. 11. National Research Council. 1987. Minorities: Their Underrepresentation and Career Differentials in Science and Engineering (Proceedings of a Workshop). Washington, D.C.: National Academy Press. 12. Institute of Medicine. 1985. Minority Access to Research Careers: An Evaluation of the Honors Undergraduate Research Gaining Program. Washington, D.C.: National Academy Press. 13. NSF Science and Engineering Education Sector Studies Group. 1988. Selected Data on Graduate Science/Engineenng Students and Postdoctorates, Fall 1987. Washington, D.C.: National Science Foundation. 14. National Science Foundation. 1988. Doctoral Scientists and Engineers: A Decade of Change. NSF 88-302. Washington, D.C.
138 FUNDING HEALTH SCIENCES RESEARCH National Research Council. 1989. Summary Report 1987, Doctorate Recipients from United States Universities. Washington, O.C.: National Academy Press. 16. National Research Council. 1987. Women: Their Underrepresentation and Career Differentials in Science and Engineering (Proceedings of a Workshop). Washington, D.C.: National Academy Press. National Science Foundation. 1988. Characteristics of Doctoral Scientists and Engineers in the United States: 1987. NSF 88-331. Washington, D.C 18. NSF Division of Policy Research and Analysis. 1989. Future Scarcities of Scientists and Engineem: Problems and Solutions. Washington, D.C.: National Science Foundation. (Working draft; April 25,1989.) 19. National Research Council. 1989. Biomedical and Behavioral Research Scientists: Their Raining and Supply. Washington, D.C.: National Academy Press. 20. Idckman, H., S. Coyle, and Y. Bae. 1990. On Time to the Doctorate: A Study of the Increased Time to Complete Doctorates in Science and Engineering. Washington, D.C.: National Academy Press. 21. American Medical Association. 1987. Physician Characteristics and Distribution in the U.S. Chicago. 2Z Jonas, H.S., and S.I. Etzel. 1988. Undergraduate medical education. JAMA 260~8~:1063 1071. 23. Jonas, H.S., S.I. Etzel, and B. Barzansky. 1989. Undergraduate medical education. JAMA 262(8):1011-1019. 24. Institute of Medicine. 1988. Resources for Clinical Investigation. Washington, D.C.: National Academy Press. 25. National Science Foundation. 1987. Report on Funding [lends and Balance of Activities: National Science Foundation 1951-1988. NSF 88-3. Washington, D.C. 26. Institute of Medicine. 1985. Personnel Needs and Training for Biomedical and Behavioral Research. Washington, D.C.: National Academy Press. 27. U.S. Department of Health and Human Services. 1989. NIH Data Book 1989. National Institutes of Health Publication No. 9~1261. Bethesda, Md. 28. Moskowitz, Jay, Associate Director for Science Policy and Legislation, National Insti tutes of Health. 1989. Presentation at AAAS Symposia on Research and Development in the FRY 1990 Federal Budget. Washington, D.C. April 1989. 29. Alcohol, Drug Abuse and Mental Health Administration. 1988. ADAM HA Data Source Book, FY 1987. Program Analysis Report No. 88-12. Washington, D.C.: U.S. Department of Health and Human Services. 30. Alcohol, Drug Abuse and Mental Health Administration. 1989. ADAMHA NRSA Research Training Tables, FY 1988. Program Analysis Report No. 89-15. Washington, D.C.: U.S. Department of Health and Human Services. 31. The Howard Hughes Medical Institute. 1987. Annual Report for 1987. Bethesda, Md. 3Z The Robert Wood Johnson Foundation. 1987. Special Report: The Foundation's Minority Medical Training Programs. Number 1. Princeton, NJ.
