Click for next page ( 141


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 140
6 Training and Career Structures in Biomedical Research Research at academic institutions traditionally has two primary ob- jectives: to increase knowledge for the public good and to train the next generation of scientists. Thus, much of the research un- dertaken in academic laboratories is performed by "scientists in train- ing," also known as graduate students, postdoctoral fellows, and clinical fellows. These trainees are expected to learn the essential skills of their field through direct experience designing, conducting, and analyzing ex- periments or clinical studies under the supervision of an established scien- tist who serves as their mentor. Scientists who emerge from this training period and wish to follow the traditional academic career path of their mentors must obtain a tenure-track position and prove their ability to es- tablish and maintain an independent research program, most often judged by their publication record and the amount of funding they obtain. The current system is well entrenched, and is designed primarily to produce new academic scientists to follow in their mentors' footsteps. As science changes, however, it may not always be the optimal approach for meeting the dual objectives of academic research. In fact, recent data indicate that only a minority of Ph.D. scientists establish a tenure-track career in aca- demia (National Research Council, 1998b, 2000, 2001a). Many Ph.D. scien- tists work in non-tenure-track academic positions, in research positions in industry or government laboratories, or in other types of science-related jobs (see Figure 6-1 and 6-2~. The recent emergence of large-scale projects in the biomedical sciences, in particular, could present significant chal- lenges in meeting the needs of trainees and junior scientists, as the tradi- 140

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH A. One to three years since doctorate Academic Track 1 4,000 - 1 2,000 - 1 0,000 - 8,000 - 6,000 4,000 - 2,000 - o 1 0,000 - 8,000 6,000 Q Z 4,000 2,000 O- Faculty . ~ \ ~ . , . . . . . . . . . 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Year ~ Total on tenure-track Type of Appointment ~ Total on nontenure-track T f f T f T f f T T T f T 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Year Postdoctoral Scientist - - - Other 141 FIGURE 6-1 Number of recent science and engineering Ph.D.s employed in aca- demia by type of appointment and academic track, 1973-1999. A: One to three years since doctorate. B: Four to seven years since doctorate. SOURCE: National Science Foundation, 2002, Appendix Table 5-27.

OCR for page 140
142 LARGE-SCALE BIOMEDICAL SCIENCE B. Four to seven years since doctorate Academic Track 1 2,000 - 1 0,000 - ...................................................................................................................................................... 8,000 z 6,000 4,000 2,000 o Total on tenure-track Type of Appointment 1 2,000 1 ................................................. ma/ ~_~ hi.. 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Year ~ Total on nontenure-track 1 0,000 8,000 - 6,000- 4,000 2,000 O Faculty \ . ~ 1 1 1 1 1 1 1 1 1 1 1 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Year _ Postdoctoral scientist - Other

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH 1 00,000 - 80,000 Q he 60,000 40,000 20,000 o 143 ~ 1 Academic ~ Industrial ~ Government 1993 1995 1997 1999 Year FIGURE 6-2 Life science Ph.D.s by employment sector. SOURCE: National Science Foundation, 2002, Appendix Table 3-18. tional structure of training and career paths in academic research may be at odds with efficient and effective endeavors in large-scale science. The effects on the career trajectories of biomedical scientists of work- ing within large-scale projects have not been studied, but by examining career issues more broadly in the field, it may be possible to identify potential obstacles faced by such scientists. This chapter provides an over- view of the current system and the challenges it presents to scientists in general and to biomedical scientists in particular. THE TRADITIONAL ACADEMIC TRAINING AND CAREER STRUCTURE IN BIOMEDICAL SCIENCE The hierarchy of academic degrees dates back to the universities of thirteenth-century Europe, which had faculties organized into guilds (see Box 6-1~. Degrees1 were in effect professional certifications that faculty members had attained the guild status of a "master." The traditional aca- 1 There was originally only one degree in European higher education, with the equivalent titles of either master or doctor.

OCR for page 140
44 LARGE-SCALE BIOMEDICAL SCIENCE

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH 145 demic training experience in the United States can also be likened in some ways to the medieval European apprentice model of craft guilds, in which a master craftsman agreed to instruct a young person in exchange for shel- ter, food, clothing, and medical care. The apprentice would bind himself to work for the master for a given time. After the agreed-upon time period he would become a journeyman, working for a master for wages, or set up as a master himself. Similarly, a student interested in pursuing an advanced degree in the sciences most often enters as an "apprentice" or graduate student in an academic research laboratory and in a sense "binds" his or her training to one specific scientific mentor or "master." In return, the gradu- ate student receives a stipend to cover basic living expenses and is offered intense training in the scientific discipline of the mentor. This training, which lasts a median of 7-8 years (National Research Council, 1998b), ide- ally integrates a broad array of professional activities, such as conducting research, writing articles and grants, and teaching. Compared with the medieval guilds that supervised the relation of master and apprentice and monitored the number of apprentices in a given guild, however, current arrangements between graduate students and their mentors in the life sci- ences are much less monitored or regulated. Upon completion of graduate training, the next step on the traditional academic career track in the life sciences is most commonly a postdoctoral position similar to that of a "journeyman" in which scholars work on a full-time but temporary basis to gain additional research experience in preparation for a professional research career. The roots of postdoctoral training in the United States date back more than a century to the 1870s, when high-level apprenticeships became part of new research institutions modeled after European examples. The postdoctoral fellowship is essen- tially a hybrid of the German privatdocent and the English fellowship, in which scientists acquire skills, prove themselves as independent scientists, and seek faculty openings (Hackett, 1987~. The Johns Hopkins University adopted the apprenticeship model shortly after its founding in 1876, and in the 1920s the Rockefeller Foundation established formal postdoctoral fel- lowships in physical science, recognizing that physics had become too com- plex to learn within the time limits of traditional programs. After a postdoctoral fellowship, the next step in a customary aca- demic career path is securing a tenure-track faculty position (National Science Foundation, 2002~. Traditionally, tenure has guaranteed the per- manence of a faculty position awarded upon successful completion of a probationary period, usually 7 years. Tenure was designed to enforce academic freedom and to make an academic career more attractive by providing job security. In practice, however, tenure in the biomedical

OCR for page 140
146 LARGE-SCALE BIOMEDICAL SCIENCE sciences provides minimal job security because it does not come with a guarantee of laboratory space or money for research. The tenure process is usually quite rigorous and can sometimes span the entire career of a faculty member. The usual professorial series, which forms the core of most academic faculty, consists of three ranks: assistant professor, associate professor, and full professor. Each rank is divided into steps of a review process. Professorial advancement up the ladder is not guaranteed and is commonly based on a multilevel merit review sys- tem that varies from one institution to another. Faculty members are usu- ally evaluated on four main criteria: research and creative work, profes- sional competence and activity, teaching, and university and public service. In the life sciences, assessment of the first two criteria is most critical, and generally entails a review of the professor's publication record and level of funding. The multilevel review system most often involves contributions from the individual, the department, the dean, academic senate committees, and the chancellor or vice chancellor. This multilevel procedure is intended to ensure that colleagues and administrators evalu- ate the professional achievements of the individual in a balanced way and in accordance with clearly defined procedures. An example of a tenure review process, used by the University of California system, is described in Box 6-2. In this case, new assistant pro- fessors are appointed initially to a 2-year contract and have a maximum of 8 years to demonstrate excellence as scholars and earn tenure. At 2-year intervals during this probationary period, candidates are evaluated and informed of their strengths and weaknesses. A midcareer appraisal, gen- erally during the fourth year, is the major performance evaluation before the tenure review. The tenure review, which generally occurs in the sixth or seventh year, leads to the chancellor's final decision on whether to grant tenure. Both the midcareer appraisal and the tenure review include the solicitation of evaluations from experts outside the university. While the granting of tenure may be the most important decision affecting a faculty member's career, the post-tenure merit-based review system has recently entered the academic tenure structure at many insti- tutions and has become a hotly debated national issue, especially as it pertains to the philosophy of academic freedom. At several institutions, it is common for an associate professor to be reviewed every 2 years, and after approximately 6 years to undergo a full review, similar in complex- ity to the tenure review, for promotion to full professor. The intent of this continuing review process is to maintain the excellence of the faculty and to reward faculty members on the basis of merit (Box 6-2~.

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH 147

OCR for page 140
148 LARGE-SCALE BIOMEDICAL SCIENCE OVERVIEW OF TRENDS IN THE BIOSCIENCE WORKFORCE Ph.D. Scientists Over the last several decades, dramatic shifts have occurred in the U.S. scientific workforce. The number of Ph.D.'s awarded annually in the life sciences has more than tripled since the 1960s (National Research Council, 1998b; see Figure 6-3~. In fact, almost all of the growth in Ph.D. production in the United States in the last several decades has been in the biomedical fields. In the last 15 years, an increasing percentage of these degrees have been awarded to citizens of other nations, more than half of whom choose to remain in this country (National Research Council, 1998b; National Science Foundation, 2000; see Figure 6-4~. During this same pe- riod, tenure-track faculty positions have declined sharply, and the num- ber of full-time non-faculty or non-tenure-track positions has increased (National Science Foundation, 2000; Hackett, 1987~. The shift to such posi- tions could potentially have a negative impact on the field as a whole in a variety of ways, including reduced autonomy and security for early-ca- reer scientists (see Box 6-3~. Two net results of these combined changes are an increase in the length of postdoctoral training and a marked reduc- tion in the opportunity for Ph.D. scientists to obtain a tenure-track faculty position. Ph.D. scientists in research universities are now more likely to be in non-tenure-track positions than to have tenure-track faculty ap- pointments (National Science Foundation, 2002; see Figure 6-1~. In the life sciences, which account for about two-thirds of all postdoctoral scientists in the United States, a postdoctoral appointment can last 3 to 5 years or 8,000 - 6,000- tn ~ 4,000- o ~ 2,000- he O- Total Biomedical ~ Nonbiomedical ~ 1963 1971 1988 Year 1 995 2000 FIGURE 6-3 Number of U.S. life science Ph.D.s awarded annually, by broad field, 1963-2000. SOURCE: National Research Council, 1998; National Science Foundation, 2002.

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH 1 2,000 1 0,000 8,000 Al ~ 6,000 at 4,000 2,000 o 149 . ~ . i. - ,, i. - . i' ~1 .. .. ~ Foreign natural science doctoral recipients ...... With plans to stay ~ With firm plans to stay 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Year FIGURE 6-4 Foreign doctoral recipients in natural science with plans to stay in the United States, 1990-1999. SOURCE: National Science Foundation, 2002, Appendix Table 2-32. more, and it is not unusual for an individual to assume multiple consecu- tive postdoctoral positions. Concomitantly, the average age of scientists receiving NIH grant support has increased dramatically. The percentage of grants going to scientists under age 35 has declined steadily from 23 percent in 1980 to 4 percent in 2001 while the percentage going to scien- tists aged 46 and older has grown sharply from about 20 percent to about 60 percent (Goldman and Marshall, 2002~. Although other career options besides academic tenure-track posi- tions exist for Ph.D. scientists (e.g., in industry, funding agencies, patent law, science writing), a recent survey indicates that the majority of doc- toral students enter their graduate program with an interest in a faculty career (Golde and Dore, 2001~. Thus there is intense competition for the academic positions available, similar to that seen in a tournament (Free- man et al., 2001~. Like an Olympic sporting event, a tournament amplifies small differences in effort, ability, or productivity into large differences in recognition and reward, intensifying the competition and creating a dis- proportionate incentive to win (Lazear and Rosen, 1981~. In the case of biomedical research, an independent career in academia might represent

OCR for page 140
150 LARGE-SCALE BIOMEDICAL SCIENCE the "prize." In an ideal setting, tournament job markets can be socially efficient, inducing high productivity from all participants. Given the cur- rent career structure of biomedical science, however, the tournament mar- ket incentives involved tend to benefit senior investigators at the expense of new entrants (Freeman et al., 2001~. Postdoctoral scientists with foreign citizenship may be especially vul- nerable within this system. The number of foreign nationals taking post- doctoral positions in the United States has quadrupled since the mid- 1970s (National Research Council, 1998b, 2001a). In recent years, half of all postdoctoral scientists in academia and in NIH intramural laboratories have been foreign citizens (National Research Council, 1998, 2000~. The status of these postdoctoral scientists is determined to a large extent by

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH 151 their visa. The most common visa options2 the "I" student visa and the "H" professional visa both have substantial drawbacks when applied to postdoctoral scientists (National Research Council, 2000~. For example, NIH training grants cannot support foreigners on student visas. In addi- tion, foreign nationals on a I visa commonly depend on their advisors for visa extensions or conversion to a green card, creating the potential for abuse. To complicate matters, when mentoring problems arise, foreign postdoctoral scientists may be restricted from changing advisers. Some data indicate that foreign nationals in the United States can compete well for positions beyond the postdoctoral level. Data collected by the Association of American Medical Colleges indicate that nearly one- third of new hires of Ph.D.s and M.D.s in basic science departments in the late 1980s and in the 1990s were foreign nationals (National Research Council, 1998b). However, making the transition from a postdoctoral po- sition to more permanent employment can be difficult because of visa issues. For example, the H work visa, which has a time limit of 6 years, requires that a petition be filed with the Immigration and Naturalization Service by a company or organization in the United States, and the appli- cation is filed for positions rather than for particular individuals. The supply of Ph.D.s in the life sciences in the United States is inevi- tably linked to the demand for work in laboratories because of the inter- connectedness of training and work in the field (Freeman et al., 2001~. Postdoctoral scholars and graduate students make economical and effec- tive workers in the laboratory because they are motivated by the hope of achieving an independent research career and making important contri- butions, rather than by monetary incentives. Thus the performance of research in the United States has relied more and more on Ph.D. scientists in training or non-tenure-track positions. As a whole, this portion of the workforce has become indispensable to the scientific enterprise, perform- ing a substantial portion of the nation's research. In many laboratories, these "junior" scientists help write grant proposals and papers; present the laboratory's research results at professional society meetings; and also educate, train, and supervise other members of the laboratory (National Research Council, 1998b, 2000~. Indeed, a 1999 survey of research articles published in Science found that 43 percent of the first authors were in postdoctoral positions (Vogel, 1999~. Shifts in the Ph.D. workforce have been driven largely by increased funding for biomedical science, which has led to more grants that include money to pay for the stipends of trainee research assistants working on the funded project. Over the last 25 years, the number of graduate stu- 2 S e e .

OCR for page 140
52 LARGE-SCALE BIOMEDICAL SCIENCE TABLE 6-1 Number and Percentage of Graduate Students of Various Kinds and Sources of Support, 1975 and 1995 1975 1995 No. % of group No. % of group Federal support Research assistant 4,653 41.7 11,963 66.5 Trainee/fellow 5,944 53.6 5,391 30 Teaching assistant 118 1.1 155 0.9 Other 404 3.6 471 2.6 Total federal 11,119 100.0 17,980 100.0 Institutional support Research assistant 3,876 25.3 8,489 38.2 Trainee/fellow 2,040 13.3 4,017 18.1 Teaching assistant 8,495 55.5 8,589 38.6 Other 901 5.9 1,136 5.1 Total institutional 15,312 100.0 22,231 100.0 Other Self-supported 9,359 71.8 6,396 55.5 Private and foreign 3,676 28.2 5,124 44.5 Total other 13,035 100.0 11,520 100.0 Grand Total 39,466 100.0 51,731 100.0 SOURCE: National Research Council, 1998:26. dents supported as research assistants through federal or institutional funding awarded to principal investigators has increased greatly, while the number of students supported directly by federal training grants or fellowships has decreased (National Research Council, 1998b; see Table 6-1~. The number of postdoctoral positions supported on principal inves- tigator grants is even greater (National Science Foundation, 2002), reflect- ing the relatively heavy use of postdoctoral scientists in university-based biomedical research, as well as the growth of independent research insti- tutes, which hire postdoctoral scientists but do not train graduate stu- dents. The vast majority of postdoctoral scientists work in universities (see Figure 6-5) as research associates on principal investigator grants, but the exact number of grant-supported postdoctoral positions is unclear because different institutions use different titles to describe them. Fur- thermore, major funding agencies (e.g., the National Science Foundation [NSF] and NIH) do not have a mechanism for counting or tracking post- doctoral scientists (National Research Council, 2000~. NIH does make an effort to track graduate students and postdoctoral scientists supported by training grants and fellowships (the National Re-

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH 20,000 - 1 5,000 - 1 0,000 - 5,000 o 153 ~ 1 Academic ~ Industrial HI Government 1981 1985 1989 1993 1997 Year FIGURE 6-5 Number of postdoctoral appointments across employment sectors, 1981-1997. SOURCE: National Research Council, 2001, Table B-2. search Service Awards, or NRSA), but it currently does not track trainees supported by grants to principal investigators. NIH is experimenting with an electronic data system that could potentially track all trainees, but that tool is not likely to be widely used in the near future (see Box 6-4~. And even in the case of NRSA fellows, NIH currently does not distinguish among different scientific specialties or characterize the recipients by other variables that could indicate the size and scope of their projects. Most universities do not track the career progression of their Ph.D. graduates or postdoctoral scientists, either3 (Freeman et al., 2001~. As a result, little information is available for assessing the effects of different work envi- ronments or projects on the career trajectories of junior scientists. There is also a paucity of data available on scientists who work in 3 This is in contrast to most professional schools (e.g., law, business, medicine) associated with major universities, which track the careers of their graduates and make the resulting information available to the public.

OCR for page 140
54 LARGE-SCALE BIOMEDICAL SCIENCE other non-tenure-track positions, although they make up a substantial fraction of the scientific workforce at many universities, especially medi- cal schools (Barinaga, 2000~. Like postdoctoral positions, these positions come with titles ranging from researcher or research associate to adjunct or in-residence professor. What they have in common is that grants rather than their institutions pay the salaries of these individuals, and they have little or no long-term job security. The majority of non-tenure-track scien- tists work within collaborative groups, so most positions are under the support and supervision of a tenured faculty member. One advantage of such positions may be the opportunity to work within a well-funded premier research team doing cutting-edge science. On the other hand, these scientists may be quite limited in their ability to establish an inde- pendent reputation for the work they perform or to reap the rewards of scientific accomplishment. Some positions do come with independent in- vestigator status, but even those who have established successful inde- pendent research programs from such positions often report a feeling of "second-class citizenship" compared with tenured faculty in their institu- tions (Barinaga, 2000~.

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH M.D. Scientists 155 There are no comprehensive sources of data for examining the factors influencing the education and career outcomes of physician scientists. However, several recent studies have examined data from a variety of sources and reached similar conclusions that the number of physician scientists is not keeping pace with the recent expansion of biomedical research, leading to a scarcity of physicians trained to undertake clinical research (Zemlo et al., 2000; Heinig et al., 1999; Institute of Medicine, 1994, 2000; Nathan, 1997~. Although the number of M.D.s in clinical de- partments of medical schools did increase between 1980 and 1994, the fraction of NIH-funded researchers in clinical departments who are phy- sician scientists has declined (Zemlo et al., 2000~. Moreover, although the number of NIH grants going to M.D. applicants increased by 32 percent between 1970 and 2001, the fraction of NIH research project awards to physician scientists declined steadily until the mid-1980s and remained level since then4 (Heinig et al., 1999), despite the fact that M.D.s are as successful as Ph.D.s in securing NIH research grants (see Figure 6-6~.5 Perhaps more telling are indications that the number of first-time physi- cian applicants for NIH funding has declined sharply in recent years, and that these applicants are far less likely than Ph.D. applicants to modify and resubmit their proposals (Nathan, 1997~. As in the case of the increasing number of Ph.D. life scientists, these opposite, downward trends in the number of physician scientists are as- sociated with financial issues, but of a different sort. As teaching and research hospitals grapple with the new realities of cost containment and managed care policies, they require more patient care time from their medical faculty, leaving less time for research. These expectations, com- bined with the relatively long training period for a research career and the large debts upon completion of professional training, have apparently discouraged many new M.D.s from following a research career path, which is likely to provide a lower salary than a clinical career path (Zemlo et al., 2000; Heinig et al., 1999; Institute of Medicine, 1994, 2000; Nathan, 1997~. 4 The number of grants going to M.D. applicants increased from 1,202 to 2,839 between 1970 and 2001. The number of grants awarded to Ph.D. applicants increased by 68 percent during the same time period, from 1,960 to 6,137 (see ~. 5 Based on NIH-wide averages of success rates for competing research project grants. Paylines at NIH vary greatly both across and within Institutes and Centers, depending on the review section and type of grant. For example, in fiscal year 2002, the overall success rate for NCI grant applicants was 29 percent, but ROls were funded to the 22nd percentile, while POls were funded to the 40th percentile.

OCR for page 140
156 to . ZU,UUU - . _ Q o A. _A / 1 5,000 - 1 0,000 - 5000 50 - 40 Al to to Al ~ 30 CO 20 - 1970 1974 1978 1982 LARGE-SCALE BIOMEDICAL SCIENCE ~ M.D.s ~ Ph.D.s O l 1970 1974 1978 1982 1986 1990 1994 1998 Year B. ~ Ph.D. A 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Year 1986 1990 1994 1998 FIGURE 6-6 NIH competing research project applications by degree. A: number of applications. B: Success rate (NIH-wide averages). NOTE: M.D. category con- tains M.D./Ph.D.s. SOURCE: . Several other challenges have been suggested as obstacles to success- ful careers in clinical research in particular. These include the complexity of working with human subjects; the long timeframe involved in many studies; and the requirement for the involvement of multiple investiga- tors, which leads to difficulties in assigning credit for papers with mul- tiple authors (Shine, 1998~. At the University of California-San Francisco (UCSF), focus groups reported that, with regard to promotions, serving as an essential collaborator on a project with other principal investigators was valued less than work done under an independent NIH RO1 grant. They also noted that this difficulty was exacerbated by departmental poli- cies that discouraged members from submitting grant proposals through organized research units or functioning as coinvestigators on grants to

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH 157 members of other departments (Task Force on the Future of Clinician Scientists at UCSF, 2000~. A lack of adequate mentoring has also been identified as a possible cause for the decline of clinical scientists (Zemlo et al., 2000; Task Force on the Future of Clinician Scientists at UCSF, 2000~. POTENTIAL IMPACT OF LARGE-SCALE RESEARCH ON BIOMEDICAL TRAINING AND CAREER STRUCTURES Because large-scale projects can vary greatly in their objectives and methods, it is not possible to generalize the career- or training-related obstacles to all cases. Each project may involve unique considerations. Noted here are a variety of potential issues that may arise within large- scale projects in biomedical research. As discussed in the previous chapter, large-scale projects require dif- ferent management structures and oversight as compared with the more traditional small-scale projects. As a result, principal investigators who receive funding to undertake a large-scale project often need to hire scien- tific managers for the project. It can be difficult to recruit such people, as they must be well versed in the science and technology of the project and also have managerial skills, which are generally not taught in life science training programs. Furthermore, because of the high costs and high pro- file of large-scale projects, the expectations for such scientist managers are very high, yet adequate incentives and compensation may be lacking, both professionally and financially. Because these scientists are funded through the grant to the principal investigator of the project, they are not likely to be on the tenure-track, so they have relatively little job stability if funding for the project is cut. If they are successful in overseeing the project to completion, most of the credit is likely to go to the principal investigator, and there may be little opportunity for promotion or other types of compensation that one finds in industry where salaries for such managers are also considerably higher. Even if the project managers are on tenure-track, participation in large collaborations may not be valued as highly by review committees as more traditional, independent work. The same is true for junior faculty (M.D.s or Ph.D.s who are on tenure-track but not yet tenured) who join a project as collaborating investigators. Graduate students and postdoctoral scientists who work on large- scale endeavors could be at risk because such projects may have a time- frame that does not fit readily with normal training programs. They may invest large amounts of time and effort, only to emerge with little to show for that investment with respect to a publication record or a scientific reputation that would help them land a position on the next rung of the career ladder. Even if publications do result from the work, the papers often include many collaborative authors, so the specific contributions of

OCR for page 140
158 LARGE-SCALE BIOMEDICAL SCIENCE trainees may be difficult to ascertain. In addition, generating pools of data for a large-scale project may be very labor-intensive and can entail a great deal of extremely repetitive work. Although laboratory research in gen- eral can often be repetitive and tedious in day-to-day practice, the scale and scope of very large projects can amplify this characteristic for ex- ample, by focusing primarily on a single method. As a result, trainees may not obtain the needed variety of experience in their training when working within such projects. The Human Genome Project provides a clear example of many of these issues. As noted in Chapter 3, the project was initially organized in the traditional fashion, with principal investigators being funded to un- dertake small pieces of the overall project. Much of the work was per- formed by graduate students and postdoctoral scientists, who were ex- pected to learn and carry out all the necessary procedures for mapping and sequencing a particular region of the genome. But progress on the project was slower than expected, so there was considerable incentive to streamline the process to meet projected deadlines, particularly after Ce- lera provided competition for the public project (see Chapter 3; Sulston and Ferry, 2002~. As a result, funding was redirected toward a few major centers that were reorganized using an industrial model. Many non-Ph.D. technicians were hired to perform the repetitive work within core groups based on specific methods, and a small number of Ph.D. scientists were hired as scientific managers. Graduate students and postdoctoral scien- tists were largely removed from the data-gathering aspects of the project.6 When the draft sequence of the public project was published in Nature, the paper included more than 250 names (Lander et al., 2001~.7 Whether all of these authors met the usual academic standards for authorship is not known, but even if only a fraction of them did, the author list would still be considerably longer than is typical of a paper in the biosciences. The fate of the graduate students and postdoctoral scientists who worked on the Human Genome Project in the early years of the project is unknown because they have not been tracked. One might perhaps predict that such individuals would have fared better than the average life scien- tist, since the field of genomics is very new and expanded rapidly in conjunction with the project. In fact, individuals with training in bioin- formatics were in short supply and highly sought after as the field of genomics expanded (Stephen and Black, 2001~. A review of available po- sitions in genomics, proteomics, structural biology, and bioinformatics advertised in two randomly chosen issues of Science magazine in 2001 6 Trainees with an interest in bioinformatics were still involved in data analysis. 7 The paper published by the Celera team also included more than 250 authors.

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH 80 - 60 40 o - o 0 20 o 159 C:1 Tenure-track ~ Nontenure-track Traditional Genomics-related FIGURE 6-7 Percent of tenure and nontenure positions requiring a Ph.D. adver- tised in two randomly chosen volumes of Science (Vol. 292 [5523], Vol. 293 [5537~. NOTE: Traditional positions include listings for cell biology, molecular biology, and biochemistry. Genomics-related positions include genomics, proteomics, bio- informatics, and structural biology listings. actually showed a high percentage of tenure-track positions (see Figure 6- 7) compared with the more traditional fields of biochemistry, cell biology, and molecular biology. As the former fields mature, however, this phe- nomenon may not continue. As more tenure-track faculty positions are filled and these scientists obtain funding, the number of graduate stu- dents and postdoctoral scientists in the field will increase significantly, thereby increasing the competition for future positions. An increase in the size of the field could also make it more difficult for junior scientists to establish a reputation based on work performed within a project rather than on the number of publications with primary authorship. It is just that sort of reputation for personal scientific achievement that has always been critical for scientists in such fields as high-energy physics, in which pub- lications include many authors listed in alphabetical order, but the field is small enough for the main players to recognize the relative contributions of the various participants. Even in that field, however, postdoctoral fel- lows and junior scientists have struggled to establish careers in their dis- cipline, dealing with many of the same issues described here for life scien- tists (Glanz, 1998; American Institute of Physics, 1992~.

OCR for page 140
160 LARGE-SCALE BIOMEDICAL SCIENCE SUMMARY Little effort has been devoted to studying the impact of large-scale research on the career paths of life scientists. Without such an analysis, it is difficult to determine whether changes are needed. This lack of assess- ment is due in large part to a dearth of outcome data for trainees in biomedical science. NIH could fill this gap by expanding its capacity to track the career trajectories of trainees supported by any type of NIH funds and to assess their training environment. Despite the paucity of such information, a number of potential ob- stacles in the career paths of scientists associated with large-scale research can be postulated now by examining trends in the field as a whole. Life scientists, both M.D.s and Ph.D.s, are struggling to establish traditional academic tenure-track careers in biomedical research, albeit for different reasons. Although a variety of other career options, including jobs in industry, are available to scientists, academic training programs generally are not geared toward those career paths, and students may find it diffi- cult to obtain the information and mentoring needed to move in those directions. There are also a growing number of non-tenure-track posi- tions in academia, but the traditional academic culture often does not provide the same degree of respect or compensation to scientists in these positions. In the case of M.D. scientists, the increasing demands of clinical practice and the burden of indebtedness from medical training are push- ing would-be researchers into other career paths. If large-scale projects are deemed worthy of substantial sums of fed- eral support, they also clearly warrant the highest-caliber scientists and staff to perform and oversee the work. But if qualified scientists are ex- pected to participate in such undertakings, they must have sufficient in- centives to take on the risks and responsibilities involved. Trainees need adequate breadth of training and ample opportunity to establish a reputa- tion for scientific achievement. Scientists in more senior positions need adequate remuneration, respect, and recognition for their work, whether it is independent or collaborative in nature. M.D.s need protected time to establish a successful career in research. Universities and NIH need work together to address these issues if large-scale projects are to be considered a valued component of the federal research portfolio. One way to protect trainees from getting lost in the shuffle of a large- scale project is through funding-agency policies that regulate the staffing of the projects they support. For example, NIH could require principal investigators to describe the training value of project proposals that in- clude trainee stipends in the budget. Indeed, NIH has already used this approach. The National Institute of General Medical Sciences (NIGMS) Protein Structure Initiative, which involves extensive data collection with

OCR for page 140
TRAINING AND CAREER STRUCTURES IN BIOMEDICAL RESEARCH 16 limited hypothesis-driven research, strongly discourages the use of gradu- ate students and postdoctoral scientists by requiring applicants to justify requests for their involvement and salary support. Instead, applicants are expected to include salary support for project managers and technicians. NIH and universities could also do more to provide incentives for scientists who choose to work on large-scale collaborative projects. For example, NIH could be more flexible with regard to providing competi- tive salaries for scientific managers of large-scale projects. NIH could also develop more-detailed policies regarding publication and authorship for large-scale projects undertaken with its support. Universities could revise their policies on tenure and promotion to recognize the value of contribu- tions made to collaborative projects. To accomplish this, academia may need to define new faculty and staff categories that better reflect the di- versity in the types of research now being undertaken, along with appro- priate criteria for performance evaluation and promotion. Universities could also expand their training programs to include such topics as man- agement training for students and postdoctoral scientists who plan to work on large-scale projects. Given the changing nature of biomedical science, collaborative endeavors are likely to become more commonplace. Thus, it would be very beneficial to the field to nurture young scientists who wish to take these positions and learn the necessary skills to manage such projects.