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.
Chapter 5 RECOMMENDATIONS LEVELS OF NRSA SUPPORT Basic Biomedical Sciences The key findings are as follows: The number of biomedical Ph.D. awards and labor demand (job openings) have been in approximate balance since the early 1980s; the ratio of job openings- to new Ph.D.s is expected to increase through the year 2000. 2. Industry is and will continue to be the dynamic sector for employment growth with the majority of new vacancies occurring in this sector. 3. The growth rate for research and development (R&D) employment will exceed that for total employment for the foreseeable future. This is due in large part to the relative growth of the industrial sector, whose biomedical Ph.D.s are heavily involved in R&D. Women are a growing proportion of the biomedical labor force. Given that women are less likely than men to be involved in full-time science, their increasing participation may lead to higher outmigration and less effective labor supply. Based on these considerations, the committee recommends shifting the NRSA program towards Ph.D. production. This could~b-e done in three ways: Increase the level of predoctoral support. There is evidence that enrollments in the biomedical sciences are very responsive to increasing student support.) The committee recommends that the level of predoctoral support be increased to 5,200 from its current level of 3,681 full-time equivalent positions. Improve the graduation rate of Ph.D. candidates. This would have a substantial impact on labor supply. Decrease the time needed to complete the doctorate.2 There is evidence to suggest that student support during the thesis considerably shortens this time. iIt has been estimated that first-year biomedical Ph.D. enrollment increases 1.35 percent for each 1-percent increase in the number of students who have support. See Joe G. Baker, "The Ph.D. Supply Crisis: A Look at the Biomedical Sciences, paper given at the Western Economics Association Meetings, June 21, l9X9, Lake Tahoe, Nevada. Utile increasing time to the doctorate has hampered PheD. labor supply in two ways. First, it has retarded the labor supply response to such stimuli as increasing wages and student support. Second, it has increased the investment costs of graduate education, resulting in lower enrollment. See Joe G. Baker, op. Cit. 79
The committee suggests that NIH examine the potential utility of a program tailored specifically for student support during the thesis-writing stage. Needed levels of postdoctoral training support will depend upon (1) the requirements of industry for this level of training and (2) decisions as to whether the public sector should support postdoctorals for industry. If the demand/supply balance that existed during the early 1980s is to be maintained, total postdoctoral support must increase from its current level of approximately 8,200 to around 11,500. The current level of NRSA postdoctoral support, around 3,700, should be increased gradually as the number of Ph.D. degrees increases. Serious questions have been raised about whether postdoctoral training represents a useful added educational experience designed to enhance productivity, or whether instead it is merely a holding tank for those doctoral scientists who are unable to land a suitable job.3 Evidence on this issue is sufficiently flawed by methodological problems as to leave it an open question about which convincing research is needed. A suggestive fact revealed in a survey carried out for the committee is that industrial employers of biomedical scientists prefer those with postdoctoral training credentials. The committee believes strongly that postdoctoral training programs in~biomedical fields should be retained and even expanded moderately, but recommends thorough program evaluations at both predoctoral" and postdoctoral levels. Behavioral Sciences The following are the key findings: 1. Most of the employment of nonclinical psychologists and other behavioral scientists occurs in colleges and universities. Given the assumption of modest enrollment growth in behavioral sciences, growth in demand for academic behavioral scientists will also be modest. 2. The labor market for nonclinical psychologists and other behavioral scientists has been fairly stable through the 1980s. 3. There is potential for behavioral science demand growth to exceed new Ph.D. supply in the post-1995 period. This is especially true if degree awards continue to fall. Behaviorally-based health problems (AIDS, drug and alcohol addiction, tobacco, cancer, etc.) are increasing in importance. For this reason, it is important to the nation s health that NIH/ADAMHA continue to support behavioral science research and research training. Given these considerations, the committee recommends that predoctoral and postdoctoral support be kept at their current levels of approximately 500 and 420 full-time equivalent positions, respectively. However, the committee also recommends moving support away from clinical psychology toward nonclinical psychology and other behavioral sciences. The~level of support for scholars in the area of health services research should be increased (see below). 3P. E. Coggeshall, et al., Changing Postdoctoral Career Patterns for Biomedical Scientists, Science, vol. 202 (November 3, 1978), pp. 487-493. 80
Clinical Sciences 2. The committee recommends that the current number of trainees and fellows for M:D. investigator training remain essentially as it has been (approximately 2,150~. Although the demand for M.D. investigators likely will increase, the committee feels that proposed studies of the training mechanisms and training outcomes for the various programs should be carried out before significant changes are made. It is hoped that these steps will be taken soon. The level of support for M:D. research training in the area of health services research should be moderately increased. DATA IMPROVEMENTS, NEEDED RESEARCH, PROGRAM EVALUATION, AND ADMINISTRATIVE INFRASTRUCTURE The committee has identified new activities in each of these areas that wouict improve understanding of scientific personnel issues and improve the quality of future versions of this report. The committee recommends that NIH/ADAMHA put in place the necessary interim staff and begin immediately the programs of data improvement, research, and evaluation proposed below. Small-Scale Data Improvements (not in priority order) 1. Improve employment specialty information in the Survey of Doctorate Recipients fSDR): An important component of the labor market Is occupational mobility;- this mechanism is a source of both labor supply and demand for new scientists in the biomedical and behavioral fields. It's the opinion of the committee that the list of occupations included for self-identification in the existing SDR survey form is too restrictive and makes it difficult for a Ph.D. who has moved out of science to respond to the survey. As a consequence, it is likely that outmobility from science is understated. This issue needs to be examined with a goal of improving the employment specialty information on the SDR. Add a salary item in the Doctorate Records File (DRE): The DRF is a census of all Ph.D. recipients from U.S. universities, collected continuously as degrees are granted. These data provide information on the new doctorate recipient's plans after graduation (approximately 60 percent of Ph.D.s have definite postdoctoral plans at the time the DRF survey form is completed). The inclusion of an item on expected salary for those with definite plans would give a mechanism for continuous monitoring of the relative tightness of the Ph.D. labor market by detailed fields, quality of degree-granting institution, sector of employment (academic, industry, federal lab, etc.), and other variables. These salary data would provide a baseline for analysis and modeling of the Ph.D. labor market at minimal cost. Improve response rate and evaluate nonresponse bias of the sDR4: The SDR is the main data set used in the analysis of the Ph.D. labor market. These data are based upon a biennial survey of approximately 10 percent of the U.S. doctorate work force. The survey is plagued with relatively low response 4This recommendation is consistent with that contained in a recent evaluation of the NSF data system. Se-e C. F.-Citro and G. Kolton (eds.), Surveying the Nation's Scientists and Engineers: A Data System for the l99Os, Washington, D.C.: National Academy Press, 1989. 81
rates and the attendant possibility of serious nonresponse bias. Two steps need to be taken: a. b. Analyze the extent of nonresponse bias. The last such study of these data was conducted in 1979. The committee endorses the recent effort by NSF to update this study, now under way. Assemble an ad hoc committee of experts in survey research methods to design approaches that will improve the SDR's response rate. 4. Include a postdoctoral identifier in the SDR: The evaluation of postdoctoral study as a research training method is critical if future policy is to be suggested. The current structure of the SDR makes it difficult to determine whether a scientist has participated in postdoctoral study; the inclusion of a question regarding this activity would allow for a more thorough evaluation of postdoctoral support. Larger-Scale Data Improvements (in priority order) 1. Establish a program evaluation data matrix: To assess the numerical adequacy of the nation's biomedical and behavioral research personnel and to make judgments about the quality of their training, we need both quantitative and qualitative information. The data matrix discussed in the appendix could provide such information. The committee recommends that the trainingSdata be organized and analyzed as suggested by that data matrix framework. 2. Reconcile SDR postdoctorate estimates and those of the Survey of Graduate Science and Engineering Students and Postdoctorates fGSESPJ: There are two sources of information on the level of postdoctoral utilization: the NSF/NIH SDR and the NSF/NIH GSESP. These surveys are inconsistent. The ambiguity of these data made it difficult for the committee to determine the historical trends in postdoctoral appointments; this, in turn, made recommendations about future levels difficult. This issue needs to be resolved to improve the analysis and evaluation of the NIH/ADAMHA postdoctoral effort. The effort may require a separate one-time survey. Improve research iclentification on the NIH/Association of American Medical Colleges fAAMC) Faculty Roster: Historically, the main source for data on physician/scientists has been the AAMC Faculty Roster. Subsequent studies have shown that Faculty Roster data are inadequate to define precisely the actual research activities of faculty and the quality of their research effort. A better identification of researchers and their activities must be achieved, in terms of both the AAMC questionnaires to deans of schools of medicine (Faculty Roster) and the NIH tracking of its institutional National Research Service Award (NRSA) training grant awards. Simple identification of self- designated researchers without any effort to quantify the amount and quality of research is meaningless. 5This recommendation is consistent with that of the Task Forces for the Review of NIH Biomedical Research Training Programs. See the NIH "Review of the National Institutes of Health Biomedical Research Training Programs," mimeographed, October 1989, p. xvii. 82
Needed Research (both sets of equal priority) Studies of recruitment and retention throughout the academic pipeline: We urgently need more comprehensive knowledge on recruitment to and losses from the pipeline into biomedical and behavioral science, especially about women and minorities at all stages of the pipeline. The committee recommends a program of research, beginning with one or more commissioned papers that summarize and synthesize the literature on the subject with emphasis on the evaluation of intervention programs at critical nodes in the pipeline. Given the anticipated labor shortages in biomedical sciences, improvement of graduation rates and participation in full-time science after graduation are important issues. Survey studies of former trainees with control groups: The program evaluations recommended below should contain several novel components relative to the prior evaluations described in the commissioned paper by Georgine Pion (see Volume III). One of these should be a series of surveys of former trainees and of carefully designed control groups. The surveys should concern career outcomes that are not reflected adequately in available secondary data, including job satisfaction and future plans, as well as the success of those in academe as undergraduate and graduate teachers. They should center on retrospective evaluations of their pre- and postdoctoral studies and, for former trainees and fellows, on the value of the programs. Finally, they should tap aspects of the research career that are not available from other sources, such as patterns of research collaboration (locally and at a distance) and sources of research funds other than NIH/ADAMHA. These studies can feed into the set recommended above by defining study populations as graduate school entrant cohorts--not just doctorate recipients- -who received NRSA support or who had the characteristics of a control group member. If done properly such studies could represent a major contribution, not only to needed program evaluations, but also to our understanding of the determinants and career consequences of attrition from graduate school. Recommended Program Evaluations The committee recommends two new, large-scale program evaluations that differ from prior efforts by including surveys of former trainees and fellows (see above) and field research that takes investigators onto the campuses at which the selected programs are locaters We recommend the careful selection of individual programs within the two training categories for initial evaluation. Program directors, department chairs, trainees, and other students should be interviewed. Curricula should be evaluated and aggregate survey results discussed. We recommend that the first two programs to be evaluated in this fashion be the Minority Access to Research Careers (MARC) and the Medical Scientist Training Program (MSTP). We recommend that the sequence of steps leading to these evaluations parallel those recommended for the two programs of research. Interdisciplinary Programs The committee urges NIH/ADAMHA-to continue to evolve its predoctoral and postdoctoral programs to meet changing national priorities, and we support the need for some of these programs to be interdisciplinary in nature. But it is imperative that those trained in interdisciplinary, cross-disciplinary, or multidisciplinary programs be thoroughly trained in basic disciplines and capable of rigorous work of the highest scientific quality. Recently developed programs (such as biotechnology training) can accomplish these 83
objectives, provided that peer evaluations are sustained at high levels and are based on clearly articulated and applied criteria. This recommendation is based on the committee's firm belief that the solutions to complex health problems, such as the AIDS epidemic, cannot be specified or narrowly targeted. Breakthroughs may come from studies in many basic fields and are likely to be found at the conjunctions of fields. But they will be found only by researchers whose understanding of those fields is deep and who have the flexibility of mind and breadth of training that allow such discoveries to be made. Needed Organization with Which to Implement These Recommendations We recommend that a new committee be activated no later than January 1992 in order to allow two years for the preparation of the 1993 report. If the preceding recommendations are initiated promptly, that report will require two years of effort and should represent a major contribution to our knowledge of scientific personnel. In order that the studies be undertaken and coordinated, there will have to be one or more persons responsible for their administration and possibly research involvement. Special Consideration: The Program of Study for Physician/Scientist Research Training Those committee members with experience in the training of physician/scientists believe that the program of study for physician/scientists, with emphasis on the NRSA institutional training grant mechanisms, must be radically changed to meet the health needs of the nation. Lloyd H. Smith notes that the science taught in specialty divisions of the clinical departments tends to be goal-oriented and superficial; although some physician/scientists have thrived in the environment and have had productive careers, most have been poorly prepared for sustained scholarship.6 Smith believes it is imperative for the serious physician/scientist to receive in-depth training in a scientific discipline relevant to medicine and that the- training of the serious physician/scientist should be comparable to Ph.D. programs in rigor and scope. The physician should not be burdened with clinical responsibilities during the research training period. Smith believes that at least three years of rigorous training in modern biological science is usually necessary for most individuals to achieve independence as an investigator. Those committee members who are experienced in the training of physician/scientists support Smith's proposal for such a training program and believe it should contain the following elements: 1. The training program should represent a consortium between the clinical and pre-clinical departments of the institution with joint responsibility for design and administration of the program. 2. Selection of the trainees should be made as early in the academic pipeline as possible, even during undergraduate medical education, if possible, but based on evidence of some experience and overall promise in research. Selection planning can then be coordinated for both basic clinical training and for the subsequent scientific training. The basic science departments and/or programs should participate in the selection process so that their commitment to the individual selected is ensured. 6See Lloyd G. Smith, "Training of Physician/Scientists," in Volume III of this report. 84
3. Formal coursework in the physical and biochemical sciences, as well as in the epidemiological, biostatistical, and economic disciplines,~should also be available and applied for an individual investigator as the need for that discipline arises. The committee members who are experienced in the training of physician/scientists recognize that it is important for graduates of this physician/scientist program to command a theoretical background comparable to that obtained by those with graduate degrees in these various disciplines. Clearly, the extent of required coursework must be individualized based upon the level of prior training, but it must be relevant and rigorous at the graduate level. 4. Such a training program in the sciences, above and beyond what is involved in the subspecialty training events, must be for not less than three years, and most of that must be invested in direct research experience under the direct supervision of a mentor. Completion of this training period, which may often be extended beyond the formal three-year program, would allow physician/scientists to rejoin their respective clinical department for subspecialty training in the chosen disciplines. It is recognized that some may elect to remain in basic science and will enrich those disciplines with their breadth of training and interest in human biology. Those committee members who are experienced in the training of physician/scientists suggest that over the course of a three-year training period~3modules of instruction should be specifically tailored to the needs of the physician trainee. This program of study would provide the physician/scientist with enough scientific depth to sustain a research career for a lifetime and not merely for a limited time beyond postdoctoral research training. Improved training in the fundamentals should also enable the physician/scientist to develop the flexible approach to problem-solving that is critical for successful pursuit of other lines of investigation stemming from observations made during the course of a goal-oriented research project. The committee as a whole finds merit in these proposals and recommends that NIH establish a committee, conference, or study to address the central issues concerning the program of study in postgraduate institutional training grants for physician/scientists. Special Consideration: Health Services Research Health services research aims to improve the way health care services are delivered through improved use of existing medical technology. It studies the quality, efficacy, and appropriateness of health care services as well as how these are affected by the method of reimbursement, the training of health professionals, and other aspects of the health care delivery system. 7This is consistent with survey results that indicate former postdoctoral research trainees favor changing the curriculum to include more formal coursework and less clinical medicine. See G. Levey et al., "Postdoctoral Research Training of Full-time Faculty in Academic Departments of Medicine," Annals of Internal Medicine, vol. 109, no. 5 (September 1988). These modules could include such topics as basic laboratory techniques chromatography, radioimmunoassay, protein purification, advanced instrumental techniques, fundamental principles of enzymology and molecular biology, subcellular fractionation techniques, computer technology, evaluation of experimental data, epidemiology, statistics and data base management, as well as grant and manuscript · ~ writing. ~5
Health services research is an interdisciplinary activity that requires individuals to be trained in a variety of fields, including medicine, economics, public health, sociology, statistics, psychology, and other fields in the natural and social sciences. In addition, it requires specially trained individuals who are capable of bringing disparate disciplines together to examine questions about the delivery of health services. These individuals receive interdisciplinary training in programs of health services research or public policy analysis. Those whose degree is in a field such as medicine or economics require additional training in the methods and knowledge base of health services research, which is usually received during a postdoctoral period. Little quantitative information is available about either the supply of or demand for health services researchers, although Elizabeth McGlynn's paper (see Volume III) identifies disciplines for the members of the Association of Health Services Research. The same paper also outlines a research agenda that would provide some important basic information about demand and supply. Because of the interdisciplinary nature of health services research, a study of field migration is central to an understanding of the supply of researchers in the field. How much migration into and out of the field occurs? What factors affect migration rates? How long does it take an in-migrant to become productive? How does migration affect the quality of health services research? A more mundane but nevertheless important evaluation issue is how one should identify a person, a research project, or a training program as being part of the field of health services research. Although we cannot quantify with much precision the current demand for health services researchers, there is substantial evidence that the demand will increase significantly in the near future. Members of the Congress, the Administration, and the private sector all have expressed a keen interest in obtaining answers to complex questions about controlling the quality and cost of health care. Bills introduced into both the House and Senate would authorize a wide program of health services research activities at a greatly increased level of spending beginning in 1990 (S. 702 would authorize spending $239 million over three years; H. 1692 would authorize $847 million over five years). The Administration has requested increased funding for evaluation research. Many private foundations have been funding pioneering work in this area throughout the last decade and the insurance industry recently has begun to sponsor research in this area. It seems prudent to increase funding for training in this area so that the research monies will be well spent. 86
