Click for next page ( 44


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 43
The Health Services Research Work Force To a greater degree than is true of many professions requiring postbaccalaureate education, health services researchers are defined by their work, not by their degrees. Physicians, nurses, and dentists, in contrast, are identified by their degrees, even when they are not actively practicing. Furthermore, although some health services researchers have degrees in the field of health services research, many have disciplinary degrees in public health, sociology, economics, or other fields. They may identify themselves as much with their discipline as with the field of health services research, perhaps because the field is relatively young. The health services research work force is marked by diversity along many dimensions. It includes individuals with widely varying backgrounds social scientists, behavioral scientists, statisticians, public health specialists, physicians, nurses who come to the field by many different educational paths. These paths include master's, doctoral, and clinical degrees as well as fellowships, on-thejob training, and summer or other short-term programs. Employment settings are also varied and include universities, academic health centers, government agencies, health care delivery and insurance organizations, consulting firms, and freestanding research organizations. Some organizations employ more than 100 health services researchers; others choose to contract for all or most of the research they need. Many organizations employ a wide variety of types and levels of researchers, who frequently work as teams. Such teams may include clinicians and master' e-level and doctoral-level researchers who have backgrounds in such diverse areas as internal medicine, nursing, health economics, epidemiology, and health care administration. In 43

OCR for page 43
44 / HEALTH SERVICES RESEARCH addition to working as members of multidisciplinary teams, individual researchers must, in many cases, consider the perspectives of multiple disciplines in their own particular research projects. As described in Chapter 1, the health services research work force has three components: ( 1 ) investigators who originate, design, supervise, and report basic and applied health services research; (2) researchers who assist in the conduct of health services research under the direction of others; and (3) individuals who analyze information and apply the tools of health services research in management and policy settings. The first group is composed largely of individuals with research doctorates; the latter two groups include many with master' e-level education. Depending on their level of education, experience, arid interest in health services research, clinicians may be found in all three groups. This chapter presents the results of this committee's efforts to identify and describe the work force, to relate current supply to demand, and to project future trends. The discussion illustrates the complexities created by the multidisciplinary character of the field, its relative youth, and the diversity of its membership. THE CURRENT WORK FORCE A major objective of the current study was a more comprehensive count of current health services researchers than has been available in the past. The database compiled for this report will provide others with a starting point for more detailed surveys of work force characteristics, career paths, and other important information. Methodological Issues Any effort to count and describe the health services research work force must overcome several challenges. First, because the field is multidisciplinary rather than marked by a defining degree, discipline, or credential, it is not readily tracked in yearly censuses or biennial sample surveys of individuals who have earned or are working toward doctoral degrees.' These data are a major 'The National Research Council (NRC) administers the Survey of Doctorate Recipients, a longitudinal sample survey that began in 1973. The data are collected by the NRC on behalf of the National Science Foundation. Samples are based on a yearly doctorate census that the NRC has administered since 1958. In alternate years, samples are drawn either humanities' doctorates or science and engineering doctorates. Those selected are then reinterviewed until they reach age ? die, or drop out of the survey. Although certain categories of individuals who might be health services researchers (e.g.,

OCR for page 43
THE HEALTH SERVICES RESEARCH WORK FORCE / 45 resource for those attempting to determine the supply of researchers in a field and describe their demographic characteristics, employment status, and career paths. Other databases such as those maintained by the American Medical Association also lack codes that would identify health services researchers. A second arid related difficulty is that academic researchers, a major component of most research work forces, are not concentrated in identifiable health services research programs. Rather, they are widely scattered across a variety of disciplinary departments (e.g., economics, sociology), professional schools (e.g., medicine, dentistry, business, nursing), and variably named health- related programs (e.g., health policy, health administration). Even in more traditional fields, it is "surprisingly difficult" to compile reliable information on faculty positions and openings (Browne, 1995, p. 16~. Third, the major professional organization for the field, the Association for Health Services Research (AHSR), is relatively young and small. Compared to larger and longer-established professions and disciplines, it has limited personnel and employment databases. The American Mathematical Society and the American Institute of Physics, in contrast, survey new doctoral recipients each summer and then resurvey them in the spring to determine their employment status, and the American Chemical Society recently undertook a survey to help assess unemployment and underemployment (Browne, 1995; COSEPUP, 1995~. Committee Data Collection Strategies To identify members of the health services research work force, the committee used two different strategies: one that combined existing data sources and the other that collected new information. (See Appendix A for more details on the work, which was supported by the Robert Wood Johnson Foundation and the Baxter Foundation and was carried out in early 1995 by the Wisconsin Network for Health Policy Research at the University of Wisconsin-Madison.) The first strategy was to combine lists of health services researchers from a variety of sources, most notably, the membership list of the AHSR2 and the list those awarded doctorates in public health, biostatistics, or epidemiology) are categorized, the survey is of little use for identifying new entrants to the field of health service research. NRC staff report that health services research has not met their criterion for adding a category to the survey. That criterion requires that at least ten survey recipients in each of the past three surveys identify the category in the space provided for "other" fields to be mentioned. 2This listing excludes individuals who take advantage of an employer's institutional membership, which provides AHSR benefits such as discounted registration for the annual meeting-to up to 10 individuals from the member institution.

OCR for page 43
46 / HEALTH SERVICES RESEARCH of principal investigators in HSRProj, the National Library of Medicine database of funded health services research projects. To these lists were added names from brochures of about 50 health research centers and names from other, smaller lists, including those participating in recent AHSR annual meetings and members of the Sigma Theta Tau, the nursing research honorary society. This combined database is the most comprehensive listing of health services researchers to date. It contains the following information (if available) for each individual: degree, address, name of employer, and telephone number. The second strategy involved new data collection through a survey of organizational employers of health services researchers (see Appendix A for more details). This survey was sent to nearly 500 organizations identified from AHSR records and other sources. It asked respondents to provide information about their health services researchers, their current and future research priorities, and current and future hiring. The categories of organizations that received the survey included university-based research centers, private research organizations and think tanks, research units in health plans, managed care organizations, insurance companies, pharmaceutical companies, and federal and state government agencies. Although the Wisconsin group and this committee believe that these organizations represent a large percentage of major employers of health services researchers, no data are available to document the survey's comprehensiveness or representativeness. Moreover, the response rate for the survey was only 31 percent so the committee interpreted the results with caution. To strengthen its qualitative understanding of the demand for health services researchers, the committee also conducted telephone interviews with 28 individuals representing several categories of health-related organizations including insurers, pharmaceutical companies, integrated health systems, state government agencies, and consulting firms. (Although the committee made an effort to contact a range of organizations, those participating were not a random sample, and responses may not be representative of the larger universe of organizations.) Questions focused on the type of issues currently being studied or considered for study in the next five years, the number and type of researchers employed, and experience in recruiting well-qualified candidates. Size and Characteristics of the Work Force The database that resulted from the two data collection strategies described above includes information on 4,920 health services researchers. Information on the type of degrees earned was available for just under two-thirds of those listed. Of these, 49 percent were reported as having Ph.D.s (or, in a relatively small number of cases, Sc.D.s or Ed.D.s). Clinicians accounted for about 28 percent of the group, and more than three-quarters of these were physicians.

OCR for page 43
THE HEALTH SERVICES RESEARCH WORK FORCE / 47 Geographically, 27 percent of the researchers in the database are located in the South Atlantic region of the country, primarily because of the concentration of researchers in Maryland and the District of Columbia. The next highest concentrations of health services researchers are seen in the Pacific region (16 percent) and the East North Central region (16 percent). California, with 567 health services researchers, has the largest state contingent, followed by Maryland (423), the District of Columbia (321), Massachusetts (317), and Pennsylvania (305~. As far as the committee could ascertain, only one survey of health services researchers, which was conducted in 1978 by a special National Research Council (NRC) panel, has attempted to describe the work force in more detail. That study (Ebert-Flattau and Perkoff, 1983) identified two populations of researchers. The first group, principal investigators, was older and more than 90 percent were male; more than one-third held medical degrees and two-fifths had research doctorates. The second group, former trainees (i.e., recipients of health services research traineeships), was younger and nearly one-half were female; two-thirds held research doctorates but only one-tenth held a medical degree (NRC, 1981, p. 107~. Both groups were overwhelmingly Caucasian, but the trainee group was somewhat less so (94 percent versus 97 percent). More than three-quarters of each category were engaged at some level in health services research. The percentage of time spent in research (as opposed to management or other activities) was lower than for biomedical researchers but higher than for behavioral science researchers. About 66 percent were employed in academic institutions, and about 4 percent of former trainees and 11 percent of investigators were employed in business or industry (not including health care delivery). Unemployment was very low (2 percent). Based on its familiarity with the health services research community (but lacking explicit information from the researcher database), the committee believes it likely that women are better represented in the class of principal investigators than in the 1978 survey, although they probably remain a minority. As in 1978, clinicians (especially physicians) appear to remain more common in the leadership segment of the research work force than in health services research as a whole. (For example, in 1995, more than one-third of the members of the Boards of the AHSR and the Foundation for Health Services Research (FHSR), had clinical degrees. At the association's 1995 annual meeting, 6 of the 11 individuals featured in "meet-the-expert" sessions were clinicians.) No reliable longitudinal data on the health services research work force are available. Certainly, AHSR membership, both individual and organizational, has grown from 450 individual and 25 organizational members in 1983 to 2,248 individual members and 128 organizational members in 1994. Such membership figures, however, constitute an unsatisfactory measure of overall work force size and growth for a number of reasons. Not all AHSR members are producers of health services research; some are users of this research (e.g., public and private

OCR for page 43
48 / HEALTH SERVICES RESEARCH sector administrators, students, journal editors and writers).3 In addition, some health services researchers may not hold individual membership in AHSR if they are affiliated with organizations that hold institutional memberships, and some individuals will choose not to join, perhaps because they already belong to several other professional groups. Once created, an organization can expect fast growth as it draws from an existing pool of potential members. Even later, when a greater proportion of new members will be new entrants to the field, some new members will be established researchers who have, for one reason or another, finally decided to join the organization. In sum, AHSR membership is not a good indicator of the size of the health services research work force or its growth. Supply in Relation to Demand The data Fathered by the committee did not allow quantitative assessments ~, of the relationship between the current supply ot and demand tor neaten services researchers. For example, the committee found no recent statistics on unemployment and underemployment rates, salary trends, or job vacancy rates for health services researchers. Qualitative information suggested healthy demand for researchers in academic settings (e.g., clinical departments that have added health services research to their research portfolio) and private sector organizations (e.g. health plans or insurers that have recently created or ~ _, strengthened research units). The committee also found some qualitative evidence of demand pressure in certain areas. Overall, more than half of the 154 responding organizations in the employer mail survey indicated that they have had difficulty recruiting certain types of health services researchers in the last few years. The areas most frequently mentioned were outcomes and health status measurement, health economics, biostatistics, epidemiology, and health policy analysis. (See Appendix A for a discussion of this survey and its limitations.) These results of the written survey (which as previously noted had a low response rate) were reinforced by telephone interviews conducted by the committee with representatives of health-related organizations. In these interviews, committee members heard of problems in recruiting well-trained and experienced researchers in health economics, epidemiology, database management, biostatistics, psychometrics, and organizational behavior. Those fin recent years, the AHSR has attempted to identify such individuals by asking member applicants to indicate whether or not they consider themselves health services researchers. As described in Appendix A, the database created for this study excludes those who answered "no" to this question.

OCR for page 43
THE HEALTH SERVICES RESEARCH WORK FORCE / 49 with experience in state government noted that the demand from state administrative and legislative offices was not for academically oriented researchers but for people prepared to conduct policy analyses and interpret research findings in the "real world" of policymaking. Executives in health care financing organizations mentioned that researchers with work experience in managed care or research management were hard to find. Several employers mentioned that they receive weekly calls from "headhunters" looking to recruit researchers with the backgrounds mentioned above. Some indicated that they look for people who have experience in health services and that they are generally indifferent to candidates' specific degrees. Others indicated that they looked for people who are well-trained in specific disciplines because they feel that knowledge about health services can be taught on the job. THE FUTURE WORK FORCE Estimating the Future Supply of Personnel To develop a rough estimate of the health services research pipeline, the committee combined information found in the 1991-1992 FHSR directory of graduate programs with responses to its own canvass of these programs. (See Appendix B for more details.) For the 76 respondents whose responses updated the FHSR directory information, the committee used their report of the number of students enrolled in each class. For the 45 respondents who did not return the update, the committee used the enrollment information reported in the 1991-1992 FHSR directory. (For those who responded, the committee compared the updated and earlier figures and found little change in enrollments for most programs, but the pattern for nonrespondents might be different.) The combined figures on enrollment in each class from the two data sources show approximately 1,015 master's students, 51 1 students in doctoral programs, and 197 individuals in postdoctoral fellowships. As an estimate of the health services research pipeline, this number must be interpreted cautiously. On the one hand, for the master's level and other programs that offer concentrations in areas in addition to health services research, the numbers may include students who have concentrations other than health services research. On the other hand, the numbers refer to enrollments in each class not to total enrollments.4 4By way of contrast, for medical and dental schools, enrollments are reported separately for the first through fourth years of school as well as for those graduating in a year. To obtain an estimate of total enrollment, the committee considered multiplying the enrollment figure for each program by the reported years to complete a degree two years for most master's-level programs and four to five years for doctoral programs. If

OCR for page 43
50 / HEALTH SERVICES RESEARCH Beyond attempting to identify researchers in training, projecting the future supply of researchers is a considerably more difficult task than counting current researchers in any field or discipline. For its 1994 report, the NRC Committee on National Needs for Biomedical and Behavioral Research Personnel created a parcel to examine methods for making work force projections. It critiqued the traditional demographic projection model used by previous NRC committees and concluded that the model was unsatisfactory. Specifically, the model (1) con- fused "age arid cohort effects" such that accurate projections could be provided only in a steady-state environment "when no projections are needed," (2) pro- vided no means for incorporating new entrants into a field, (3) made faulty assumptions that the ratio of students to faculty and support dollars per researcher were fixed and unaffected by changing economic and technical conditions, and (4) incorrectly assumed that a current imbalance between supply and demand would not affect the future labor market (NRC, 1994, p. 21~. The NRC panel concluded that the best approaches to work force projections involved the use of"multistate period life tables." (See Appendix C for a more detailed description.) As described in the report (NRC, 1994, p. 21), such approaches begin by listing the characteristics of a given population (e.g., age, sector of employment, and employment status) and project changes in the population based on the life his-tory of members of the population.... [Projections] are generated through a series of statistical calculations making assumptions about both the rates of transition of individuals from state to state (employed to retired, for example) and rates of new entrants to the system. This kind of analysis can help policymakers get a sense of what the characteristics of the labor force may be in the next several years. The further out the projections go, however, the less useful they become as unforeseen events affect individual choices about educational and career options. Unfortunately, despite the progress the present Institute of Medicine (IOM) committee made in enumerating the health services research work force, it still lacked the critical current and historical information identified above. The committee, therefore, chose not to attempt numerical projections of the size of the future health services research work force. It also did not attempt to calculate the number of new entrants that would be needed to maintain the supply of researchers at a particular level. Specifying a target work force size latter average were applied to the figure for doctoral enrollment reported in the text (51 1), it would yield an estimated 2,700 doctoral students in the pipeline. Based on committee members' involvement in the field, this number seemed implausibly high, presumably because it would not account for attrition.

OCR for page 43
THE HEALTlI SERVICES RESEARCH WORK FORCE / 51 requires assumptions about the future demand for workers, a process that is even more difficult and uncertain than projecting the future supply of personnel. Estimating the Future Demand for Personnel Demand is conventionally described as the willingness and ability to pay for a good or service. Need is a more subjective and often more expansive concept that may be variably judged by consumers, suppliers, or "experts." Both should be considered in any discussion of future personnel requirements. Several recent reports have expressed skepticism about "the possibility of generating useful forecasts of demand," especially long-term demand for researchers (NRC, 1994, p. 21; see also COSEPUP, 1995; and IOM, 1995~. One reason is that making estimates about key factors influencing demand (e.g., funding for research, technological developments) is a highly speculative endeavor. Demand models are "sufficiently subjective and vulnerable to changing events and data limitations that they are less useful as specific numbers than as means of illuminating supply-demand dynamics" (IOM, 1995, p. 268~. In addition, they do not adequately account for the behavioral responses of students, faculty, and others whose decisions about training, careers, and educational programs can shift supply away from projected levels (COSEPUP, 1995~. Even short-term indicators of demand are "of limited use to policymakers because today's decisions about fellowships and traineeships primarily affect the scientific labor market a decade from now" (NRC, 1994, p. 219. Such indicators, however, have some value because current market conditions are important influences on the career decisions young people make and because a tight current labor market might suggest different training support strategies. Unfortunately, the health services research field lacks easily accessible information to project short-term demand. The committee's surveys of those employing health services researchers did, however, provide some qualitative perspectives on future demand as described below. Future Prospects Although the committee did not attempt to develop a numerical estimate of the future supply of or demand for health services researchers, it did discuss what might be expected of various factors that could affect supply or demand. These factors include (1) changes in governmental and private foundation funding for higher education and research generally and for health services research and education specifically, (2) strains on academic health centers, and (3) other restructuring of the health care system.

OCR for page 43
52 / HEALTH SERVICES RESEARCH As this report was being drafted, prospects for federal funding for many programs and activities were highly uncertain. Even such traditionally popular programs as those for biomedical research and agricultural subsidies were being questioned. In the first proposals for reductions in expenditures for fiscal years 1996 to 2002, committees in both the U.S. Senate and the U.S. House of Representatives recommended elimination or major cutbacks in the Agency for Health Care Policy and Research, the lead agency for public funding of health services research and education (Brown, 19953. At the state level, the committee found concerns about the willingness and ability of many states to support and maintain ail ongoing program of health services research and sophisticated policy analysis. In telephone interviews, those with experience in state government cited civil service hiring restrictions, policymakers' skepticism about researchers' sensitivity to the policy considerations, and pressures on state budgets as reasons to be cautious about hiring at the state level. In addition, although health care reform initiatives may have increased demand for health services researchers (at the master's if not the doctoral level), several state reform initiatives face difficulties following the 1994 elections. Private foundations, although not major direct employers of researchers, can be expected to continue to support health services research undertaken by academic and other organizations. Foundations, however, are also facing heavy pressure to use their resources to offset reductions in public spending for social and educational programs, and this pressure may divert some resources from research. In many academic medical centers, economic forces and policy decisions are likely to reduce patient care income and other revenues that may have been used directly or indirectly to help support various kinds of research including some limited amounts of health services research (Epstein, 1995~. Government funding of graduate medical education and payments for services to Medicare beneficiaries are highly vulnerable to cuts. Equally serious, the relatively high costs of academic medical centers make them unattractive to managed care plans that are building provider networks. The resulting loss of patient care revenues could jeopardize the continued existence of some institutions. For other institutions, health-related research could be affected (especially in areas not of direct relevance in a competitive health care market), although cuts in external research support (e.g., government grants, industry contracts) are a greater concern. State agencies reported that civil service rules and legislative skepticism about funding research (analysis is considered a more acceptable term) made employment of researchers relatively uncommon except in a few areas such as

OCR for page 43
THE HEALTH SERVICES RESEARCH WORK FORCE / 53 epidemiology. States tend to rely heavily on consulting firms and, sometimes, university-based consultants who can provide policy-sensitive analysis on a quick-response basis. In business and industry, employment opportunities for health services researchers appear bright but still unpredictable in some respects. On the one hand, demand for data and analysis related to health care utilization, costs, and outcomes has been increasing as evidenced by the development of specialized firms and units to analyze organizational, clinical, and other information and to report on the performance of health care organizations.5 On the other hand, in a competitive market, the value of research and analysis will be routinely weighed against the value of other activities in a new decisionmaking environment that is arguably less supportive of long-term investments than that which has prevailed in nonprofit environments. In addition, economic factors may well discourage the hiring of full-time workers eligible for an array of fringe benefits and other advantages, which could mean that the episodic and leaner use of outside contractors will be favored over the use of internal staff to provide research services. The committee found consulting firms, health insurers, and similar organizations reporting that they expected to increase their hiring of master's- and doctoral-level health services researchers in the foreseeable future. Many of the employers interviewed by committee members, particularly those representing integrated health systems, pharmaceutical companies, and private consulting groups, stated they planned to double their complement of health services researchers during the next five years. In response to the written survey's request that employers predict their recruitment plans for the next five years, more than 60 percent of responding organizations indicated that they anticipated recruiting more health services researchers in at least one of the 13 identified areas of research. Thirteen of 14 responding managed care, insurance, and similar organizations indicated they expect to recruit health services researchers in the next five years. The research areas that most respondents will recruit from are outcomes and health status measurement, health economics, biostatistics, epidemiology, and health policy analysis. Based on respondents' reports about current recruiting difficulties and the committee's own experience, the committee expected that researchers with (1) skills in biostatistics, health economics, outcomes and health status measurement, and epidemiology, and (2) experience in managed care and research management would continue to be difficult to find for the next few years. Although the committee did not collect information on other characteristics that would make 5Interest in these organizations is sufficiently high that they are the subject of several directories, for example, yearly directories in the trade journal Business Insurance.

OCR for page 43
54 / HEALTH SERVICES RESEARCH individuals more attractive in the job market, it notes the conclusion of a recent study of graduate scientists and engineers: "more employment options are available to graduate scientists and engineers who have multiple disciplines, minor degrees, personal communication skills, and entrepreneurial initiative" (COSEPUP, 1995, pp. 2-17~. The implications of this discussion for training programs are considered in the next chapter. CONCLUSION The committee's success in achieving the study's objectives of describing the current size and characteristics of the health services research work force and projecting the future supply and demand for researchers was limited by data inadequacies. One priority became additional data collection to identify more completely the existing work force and, to a lesser extent, to gauge short-term demand for health services researchers. The resulting database of 4,920 health services researchers and information regarding demand proved illuminating, but a priority for others should be developing more complete information about the characteristics and employment patterns of the work force. Data limitations notwithstanding, the committee understood that the major factor affecting future employment will be public and private research funding. In the immediate future, the division of funding between these two major sources may shift somewhat toward the private side. As discussed in the next chapter, educational programs that prepare health services researchers will need to pay more attention to private organizations as potential employers of their students as well as potential funders of research.