GRADUATE-EDUCATION IN THE BIOMEDICAL SCIENCES - CRITICAL OBSERVATIONS ON TRAINING FOR RESEARCH CAREERS

Several aspects of the processes for the pre- and postdoctoral training of PhDs and the postdoctoral research training of MDs are critically examined. The size of the predoctoral pipeline, the sources of support for the students in it, trends in the annual production of bioscience PhDs, and prospects for growth in opportunities, as defined in increases in availability of public and private funds for research, are catalogued. Evidence for the existence of a surplus of research scientists-based on the size and growth of the pool of postdoctorals, the success rates now being experienced by young postdoctorals in competing for National Institutes of Health (NIH) grants, and the success of young postdoctorals in securing career employment-is evaluated, and the conclusion that the nation is producing too many research scientists is suggested. The pros and cons of ''downregulating'' the production of scientists are explored, the difficulties of reaching a national consensus on the degree of reduction and the partition of reduction among fields of science and academic institutions and academic departments are described, and the mechanisms conceivably available for accomplishing the task-reducing NIH training funds, autoregulation by academic institutions and/or scientific disciplines, and reliance on the decisions of well-informed (probably by the National Academy of Sciences) degree candidates-are enumerated; a preference of the last of these mechanisms is indicated. The NIH's formal training programs are compared with its informal support of training under research grants; questions that the latter practice raises are identified. The striking disparity between the duration of training for conventional predoctoral PhD candidates and that for dual-degree (MD-PhD) aspirants is noted. The measurement problems of assessing the duration of postdoctoral training are highlighted. The fact that dual-degree scientists seem to compete little, if any, more successfully for the great bulk of NIH research grants than do singly-degreed MDs or PhDs is noted, suggesting the advisability of a fresh and objective review of these dual-degree programs, especially the NIH's Medical Scientist Training Program, to reassess their value in the light of their cost. Some characteristics of careers in the biomedical sciences are outlined. Data are presented on the high turnover rates of first-time-ever entrants into the pool of NIH grantees or, stated otherwise, the relatively low rates of survival of principal investigators (PIs) in the NIH research grant system; a method of estimating the steady-state number of NIH ex- PIs ''around'' at any given moment is proposed. Serious problems surrounding the evaluation of training programs are discussed, leading to the surmise that no extant evaluation, including those done for the NIH by the National Academy of Sciences, the National Research Council, the Institute of Medicine, and the Association of American Medical Colleges, is entirely credible. Data systems relevant to graduate education, career outcomes, and employment patterns-both as they do and as they could exist-are discussed. A mechanism for federal support of research training, substantially different from those now in place, is described and its adoption suggested.