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 THE ACADEMIC HEALTH CENTER AS TRANSLATOR S A OF CIENCE : THE RESEARCH ROLE Advances in health care in the 20th century, especially those emerging from basic research, were remarkable. The mapping of the human genome in particular is expected to have a profound effect on the future of health care (Pober et al., 2001). Many believe that in the 21st century, the life sciences will emulate the intellectual and economic feats of the physical sciences in the last century (Greenberg, 2000). Expectations are high that the burden of disease and disability can be reduced through research. As we embark on the 21st century, it is important to maintain the capacity for continued discovery in the basic sciences. Without a greater focus on clinical, health services, and prevention research, however, the full benefits of such discoveries will not be realized. Research provides the foundation for our current scientific knowledge; the challenge in the future will be to translate that knowledge and the resulting expanded capabilities into daily practice (Frist, 2002). This chapter examines how the research role of AHCs can be used to improve the health of people and develop the scientific evidence base for health and health care. Overall, the committee finds the following: · AHCs have been significant contributors to the enormous strides made by research in the past decades, especially those of basic scientific research. Investments in basic science should be continued to support ad- vances in discovery and understanding. AHCs and their parent universities 77
78 ACADEMIC HEALTH CENTERS play a critical role in the long-term basic research that makes future innova- tions possible. · In the coming decades, continued scientific discoveries and ad- vances will require that AHCs continue their work in basic science research and discovery, as well as developing and refining the evidence base for health care by: -- Encouraging studies that embrace the continuum from animals to humans to experimental models. -- Increasing the emphasis on clinical research in order to translate new discoveries into clinical practice and evaluate current clinical practices, thereby answering questions about what does and does not work in health care. -- Increasing emphasis on health services research in order to im- prove understanding of the effectiveness and costs of care, especially the impact of new discoveries on the costs of care and treatment patterns. -- Increasing emphasis on prevention and population research in order to improve understanding of how to identify and reduce health risks, as well as the linkages between personal and population health. This chapter describes the continuum of research and the challenges that will be faced in the coming decades by both AHCs and the agencies that fund health-related research. The first section examines the processes by which the scientific evidence base for health is created and applied. The next two sections, respectively, review the continuum of research from discovery through application and the obstacles faced by AHCs in conduct- ing research across this continuum. The final section presents some implica- tions for the future. CREATING AND APPLYING THE SCIENTIFIC EVIDENCE BASE FOR HEALTH The trends and shifts described in Chapter 2 will create both opportu- nities and challenges for research in the coming decades. The rapid pace of discovery will generate opportunities to improve health in new and poten- tially more effective ways. Advances in the past have yielded great benefits in terms of outcomes of care, increased longevity, improved quality of life, and reduced absenteeism from work (Cutler and McClellan, 2001). Ill- nesses once thought to be incurable can now be treated, and sometimes cured or prevented, as a result of the scientific and technological advances made possible by basic research (Frist, 2002). Many believe that science is on the cusp of generating a major revolu- tion in medicine as a result of advances in genomics, proteomics, and such areas as stem cell biology that offer the potential for new breakthroughs in
AHC AS A TRANSLATOR OF SCIENCE: THE RESEARCH ROLE 79 tissue engineering. One researcher suggests we are "at the beginning of the end" of a phase of discovery that involves identifying the molecules required for human life (Pollard, 2002, p. 1725). Most molecular defects are seen at the cellular, organ, or organismic level. Current knowledge is linking an individual's predisposition to disease, but additional understanding of the underlying mechanisms will be needed to move toward preventive medicine at the molecular level (Pollard, 2002). It is also now feasible to conduct studies in humans that were not possible in the past, potentially improving understanding of the pathogenesis of disease. More proof-of-concept stud- ies in human subjects are needed if new diagnostic and therapeutic ap- proaches are to emerge from the laboratory to enter clinical practice. Such advances have the potential to provide powerful tools that will improve health and fundamentally alter the practice of medicine. Expectations are high that science will continue to yield great advances in the future, although the pace at which such discoveries will have a broad impact on people is unclear. The public has shown a willingness to support this important work, as evidenced by the growth in funding for the Na- tional Institutes of Health (NIH). Furthermore, concerns about bioterrorism and recurring and emerging infectious diseases will lead to more appeals for science to help alleviate and respond to such threats. The trends described in Chapter 2 will also create a serious set of challenges for research in the coming decades. Not all advances will come from great breakthroughs. Health care also advances through a slow and steady series of incremental steps that refine knowledge and technology so that, cumulatively and over time, improvements in health result. There is a need for better knowledge of how to care most appropriately for and maintain the health of an aging and chronically ill population, and how to both improve the quality of care and contain its costs. Achieving progress in these areas will require improved understanding of the effectiveness of the clinical, organizational, and financial aspects of care so that safety, effi- ciency, and effectiveness can be designed into systems of care. There is clear evidence of a gap in applying current knowledge in practice (Institute of Medicine, 2001b). Some patients are not receiving treatments that could be beneficial to them. According to one study, about 50 percent of patients for whom beta blocking agents were appropriate did not receive them (O'Connor et al., 1999). Some patients are receiving treat- ments that provide no benefit, or even cause harm. For example, calcium channel blocking agents were administered to 18 percent of patients with impaired left ventricular function, even though current guidelines recom- mend against their use in such cases (O'Connor et al., 1999). Likewise, over a 1-year period, 60 percent of Medicaid patients diagnosed with a cold filled a prescription for antibiotics (Shuster et al., 1996). Moreover, errors in clinical care result in death for thousands of people each year (Institute of
80 ACADEMIC HEALTH CENTERS Medicine, 2000). Computer-based prescription order entry systems have been shown to reduce prescribing errors by one-half to three-quarters, but it is estimated that fewer than one-fifth of hospitals currently have such systems in place, and among those that do, fewer than 10 percent of orders are computerized (Doolan and Bates, 2002). In short, there is a large gap between what we know and what we do in terms of health care. As noted in earlier chapters, unexplained variations in care have also been documented, suggesting the need for continued development of the evidence base. For example, among Medicare beneficiaries, overall dis- charge rates for medical conditions are 60 percent higher in Boston than in New Haven (Wennberg, 1999). Geographic analyses of Medicare benefi- ciaries have revealed that spending on health care in Miami was nearly 2.5 times that in Minneapolis (even after adjusting for age, sex, race, and price levels); visits to specialists in the last 6 months of life ranged from two times in Mason City, Iowa, to more than 25 times in Miami; and the proportion of eligible patients receiving beta blockers after a heart attack ranged from 5 to 92 percent (Wennberg et al., 2002). Variations in discharge rates, hospital days, and volume of outpatient visits among similar patients are found across age groups, in both inpatient and outpatient settings, for both acute and chronic conditions (Blumenthal, 1994; Ashton et al., 1999; Wennberg, 1999), and across different forms health insurance (Brook, 1997), and they persist even after controlling for differences in severity of illness. There is also growing recognition of how little is known about the effectiveness of many drugs, devices, practices, and procedures that are accepted as part of today's clinical practice (Garber, 1994), and of how difficult it is to synthesize across studies to advance knowledge. Another concern is that the rising costs of care discussed in Chapter 2 are due in part to technological advancement itself. Technology affects the costs of health care by increasing the intensity of care provided to patients and by expanding the applications of the technology and the populations who can benefit (Neumann and Sandberg, 1998). At times, an innovation is introduced while its appropriate use remains uncertain, and is refined only after being applied in practice rather than before it has been diffused (Gelijns and Rosenberg, 1994). New technologies and increased use of existing technologies have been estimated to account for as much as two-thirds of the real annual increase in health spending (Blumenthal, 2001). Although technology may improve efficiency by reducing the cost of care per person, the number of eligible patients grows over time, so overall expenditures increase as well (Weisbrod and LaMay, 1999). Therefore, cost savings that may show up at the individual patient level are offset by overall higher expenditures due to increased use of the new technology (Gelijns and Rosenberg, 1994). A concern, then, is that increased investments in bio- medical research will contribute excessively to rising health care costs
AHC AS A TRANSLATOR OF SCIENCE: THE RESEARCH ROLE 81 (Blumenthal, 2001) unless there is a better understanding of how to apply judiciously the innovations produced. All of this suggests that, despite the great advances in knowledge achieved during the 20th century, much work remains to be done in the 21st century to develop a sound evidence base for health. An improved evidence base will result in part from basic research that will continue to uncover the fundamental mechanisms of disease and thereby reduce uncer- tainty in practice. For example, some complications from treatment or adverse reactions to medications may be reduced as a result of improve- ments in basic knowledge of treatment effects and patient responses. An improved evidence base will also result, however, from an increased em- phasis on clinical, health services, and prevention research that will im- prove abilities to apply current knowledge, helping, for example, to elimi- nate recognized problems of overuse, underuse, and misuse (Chassin et al., 1998); assess the cost-efficient application of technology; or evaluate strat- egies designed to reduce health risks throughout the population. Thus AHCs need to participate in developing solutions to society's most press- ing health problems not only by creating knowledge, but also by develop- ing more systematic approaches for using research to encourage evidence- based patterns of practice, in order to improve health for both patients and populations. WORKING ACROSS THE CONTINUUM OF RESEARCH The translation of the discoveries of basic science into practice can be viewed as occurring along a continuum. This continuum has been defined in various ways, but generally progresses from basic research, to clinical research, to applied research--from fundamental science, through its appli- cation to patients, to studies of health and disease in populations (Frist, 2002; Association of American Medical Colleges, 1998). In addition, this committee considered the continuum in terms of the aim of the work-- from discovery, to testing, to application, to evaluation. Discovery tends to rely on basic research; testing and application tend to rely on clinical re- search; and evaluation tends to rely on applied research. However, these distinctions are offered as a broad framework rather than a typology. Basic biomedical research includes molecular biology, biochemistry, and cell biology and their application to mammalian, especially human, systems (Pober et al., 2001; Fontanarosa and DeAngelis, 2002). It often includes laboratory research using human material, such as cell cultures and DNA analyses (Oinonen et al., 2001). Advances in the fundamental sciences and the mechanisms of disease are critical to the development of diagnostic and therapeutic technologies and to the targeting of areas for subsequent clinical study (Gelijns and Thier, 2002).
82 ACADEMIC HEALTH CENTERS Clinical research is defined by NIH as including three areas of study: a) patient-oriented research, that is conducted with human subjects (or on material of human origin such as tissues, specimens, and cognitive phenomena) for which an investigator (or colleague) directly interacts with human subjects. This area of research includes mechanisms of hu- man disease; therapeutic interventions; clinical trials; and development of new technologies; b) epidemiologic and behavioral studies; and c) out- comes research and health services research (National Institutes of Health, 1997, www.nih.gov/news/crp/97report/esecsum.htm#2define). Although the above definition of clinical research includes aspects of study that relate to health services research, the committee has chosen to distinguish the latter from clinical research because it requires a distinct set of skills, focus, and expertise. Health services research is a multidisciplinary field of scientific investigation that studies how social factors, financing systems, organizational structures and processes, health technologies, and personal behaviors affect access to health care, the quality and cost of health care, and, ultimately, health and well-being. Its research domains are individuals, families, organizations, institutions, communities, and popula- tions (AcademyHealth, 2002). It is often considered to also include research related to health policy and management. Goldstein and Brown (1997) make an additional distinction between disease-oriented and patient-oriented research: the former is targeted to- ward understanding the pathogenesis or treatment of a disease but does not require direct contact with patients; the latter is performed by clinicians who observe, analyze, and manage individual patients. Disease-oriented research can be thought of as a bridge between basic and clinical research in that it focuses on a specific condition, as does clinical research, but it does not involve patient contact, as is the case with basic research. Goldstein and Brown perceive a rapid growth in disease-oriented research as compared with patient-oriented research for several reasons. First, technological break- throughs in molecular biology attract scientifically oriented clinicians to basic science. Second, the pressures on the health care delivery system (combined with the rapid pace of research) make it difficult for any one person to be intensely involved in both types of research simultaneously. Third, basic research is often able to produce more clear-cut results as compared with patient-oriented research, making it easier to publish the results and obtain funding. It is important to note that despite an implied order to the research process, the process by which a discovery is made, proven in practice, and diffused into the community is not necessarily linear (Gelijns and Rosenberg, 1994). For example, an innovation that enters practice usually undergoes a continuing process of refinement and development after its introduction.
AHC AS A TRANSLATOR OF SCIENCE: THE RESEARCH ROLE 83 Moreover, an experience in the clinical setting may feed back into the laboratory, leading to the development of additional fundamental knowl- edge about a condition; that is, the order of the process as outlined above may be reversed. The stages of the continuum also tend to occur in a discrete and often unrelated fashion. For example, the cost-effectiveness of a discovery is usually assessed separately from its clinical effectiveness, typically after it has been introduced into a clinical setting. Furthermore, although the continuum described above encompasses different types of research, they are all interrelated. Basic research rarely produces findings in the laboratory that can be used immediately in clinical practice. One means of translating basic research into practice is through clinical trials, but they leave many unanswered questions as well. For ex- ample, clinical trials generally do not consider impacts on overall treatment patterns for affected or multiple populations and rarely examine the costs or cost-effectiveness of an intervention (that is, they focus on efficacy more than effectiveness). Even after testing for clinical safety and effectiveness, questions remain about how to integrate the improvements into everyday practice for the average patient. Health services, outcomes, effectiveness, and other evaluative research can complement the work done in biomedical and clinical research by addressing such issues as how to organize and finance care, how to measure and evaluate its quality, how to involve patients in their care, how to encourage patients to adopt behaviors that promote health and prevent disease, and how to facilitate the adoption of scientific knowledge (Horwitz, 2002). Additionally, although the various forms of research are interrelated, they are typically conducted by different scientists and funded separately. Increased coordination and collaboration will be required to meet growing demands for rapid improvements in health care and for a greater focus on the types of research that answer questions about what does and does not work (Stryer et al., 2000). This is not to suggest that research is useful only if it has an immediate impact, but rather that the ultimate goal is to produce knowledge that can help people. AHCs have been shaped by basic research, dating back to a 1945 report by Vannevar Bush that established a system for federal support of research conducted primarily by independent investigators, based in univer- sities, and awarded funds through a process of peer review (Association of American Medical Colleges, 1998). During the 1950s and 1960s, NIH believed that diseases would be cured when science provided an under- standing of their physiology; the result was significant growth in the basic science departments of medical schools (Goldstein and Brown, 1997). Basic science is viewed as flourishing today because of growth in NIH funding levels for this work and in the number of basic researchers (Goldstein and Brown, 1997).
84 ACADEMIC HEALTH CENTERS Although almost all AHCs receive funding from NIH, this funding is concentrated in a subset of these institutions. It is estimated that in 2000, approximately one-third of NIH funding to AHCs went to the top 10 institutions (mainly to the medical schools), which received an average of about $280 million each; the next 40 institutions received about 50 percent of the money, for an average about $110 million each; and the remaining institutions received about 15 percent of the total (Anderson, 2002). Be- tween 1987 and 1997, the proportion of NIH research awards to the 10 most research-intensive institutions increased, while those to the less-inten- sive institutions decreased (Moy et al., 2000). The reliance on NIH funding could create financial constraints on AHCs' ability to maintain current levels of basic research. Sustaining the recent rate of growth in the NIH budget would require 14 to 16 percent annual increases; increases of less than 6 percent would squeeze current funding levels. The president's budget for 20042007 includes an annual growth in funding for NIH of around 2 percent (Korn, 2002). AHCs have also been affected by a shift in the way clinical trials are conducted. The volume of clinical trials is growing rapidly, and the trials are also becoming dispersed to many sites. Private companies, known as contract research organizations (CROs), manage clinical trials for pharma- ceutical companies. CROs are one mechanism for expanding the capacity to conduct trials, and they also allow physicians in the community to become involved in clinical research. At the same time, however, more clinical trials are now being conducted outside of AHCs than within them. It has been estimated that investigators in AHCs represent about 46 percent of all those involved in research, down from 80 percent a decade earlier, with the majority of industry funding for clinical trials being allocated to community-based efforts (Morin et al., 2002). The market for CROs is estimated to grow by approximately 15 to 20 percent per year, leading them to dominate the market for clinical trials research (Rettig, 2000). Obstacles to Conducting Research Across the Continuum AHCs need to consider how they can participate in research across the full continuum described above. Despite growing interest in measuring the effectiveness of medical interventions and developing more valid and robust indictors of effectiveness, AHCs face a number of obstacles in accomplish- ing these objectives. Several such obstacles are considered here, including training of clinical investigators, creation of the organizational processes required for research across the continuum, inadequate federal funding levels, and ethical issues.
AHC AS A TRANSLATOR OF SCIENCE: THE RESEARCH ROLE 85 Training of Clinical Investigators A major obstacle to conducting research across the continuum is the supply of clinical researchers. Whereas the number of basic researchers is growing, clinical advances are threatened by a lack of growth in the num- bers of clinical investigators (Goldstein and Brown, 1997). It is estimated that only about 11 percent of medical school graduates plan a career de- voted exclusively or significantly to research (Nathan, 2002). A study by the National Research Council (2000) revealed that the number of Ph.D.'s awarded in the basic biomedical sciences is well above that needed; how- ever, there is evidence of a decline in the number of physicians conducting research. Unfortunately, data for determining such trends are highly limited as no objective data source exists (Crowley and Thier, 1996; National Research Council, 2000). There are several barriers to pursuing a career in clinical research. Clinical researchers obtain training in both biomedical sciences and clinical practice, both of which are increasing in complexity. Major debts are in- curred from the many years of training required to acquire expertise in both research methods and clinical care, and the demands of retaining skills in both areas over time are enormous (Crowley and Thier, 1996; Nathan, 2002). The pressure to pay their debts causes investigators to spend more time in clinical care than in research (Wolf, 2002). Furthermore, there are fewer training opportunities in clinical, health services, and prevention re- search than in laboratory research in that the latter appears to be favored by both funding agencies (particularly NIH) and AHCs themselves (Crowley and Thier, 1996; Wolf, 2002). A lack of predictable support also raises concern about the ability to raise sufficient funds to conduct the research. This concern is particularly acute for clinical investigators who train until their mid-thirties and then may be unable to raise sufficient funds to pursue their intended work (Wolf, 2002). A lack of core institutional resources is also seen as a barrier, particularly for health services research centers (Nathan, 2002; Kindig et al., 1999). In recent years, NIH has attempted to address these concerns by creat- ing a series of awards for new and midcareer investigators involved in clinical research, as well as educational loan repayment programs (Wolf, 2002). The expansion of training opportunities at NIH-supported General Clinical Research Centers may be another approach to increase support for clinical researchers (Vaitukaitis, 2000). These centers receive support for research infrastructure, including specialized staff and computer systems, and many are located at AHCs. Clinical training programs should ensure better exposure to quality research experiences to encourage more clinicians to consider careers in research. Students exposed to research during their clinical training may be
86 ACADEMIC HEALTH CENTERS more likely to engage in research activities later in their career (Kalfoglou and Sung, 2002; Institute of Medicine, 1994). Even those students not choosing a research career can gain an understanding of scientific methods and the capabilities needed for critically evaluating the research literature (Institute of Medicine, 1994). If students are expected to read the research literature and understand the latest findings surrounding the new sciences, they need to learn the language of the field. These are regarded as valuable skills even for those who do not conduct research directly (Kalfoglou and Sung, 2002). Creation of Organizational Processes Clinical and health services research tends to be organized differently from basic science research. Basic biomedical research is typically carried out by an individual investigator or team of investigators from the same field. In contrast, the power of translational research derives from its com- bination with basic and population sciences; more interdisciplinary ap- proaches are required, as well as better integration of medical, health, social, and behavioral sciences and other areas of the life sciences. The emphasis on research performed by individual scientists may have worked well in the past but is "not a prescription for success in clinical research" (Nathan, 2002, p. 2426). Some believe that the individual investigator who tries to do it all will flounder (Nathan, 2002), given that the necessary expertise will reside in a team of researchers rather than an individual, as is becoming increasingly true for many types of research. The interdisciplinary approaches that are central to translating re- search into practice are not well rewarded either by external funders or within the AHC structure. Research studies on patients appear to be held in low esteem by NIH study sections (Nathan, 2002). Within AHCs, investi- gators engaged in patient- and population-based studies have not been promoted as rapidly as individual scientists conducting basic research and have more difficulty in obtaining discretionary resources (The Common- wealth Fund Task Force on Academic Health Centers, 2000). In evaluating individuals for promotion, AHCs have typically emphasized NIH grants to individual investigators as well as the ability to publish in leading journals, which is facilitated by having received an NIH grant. Moreover, institu- tional support within AHCs is allocated through individual departments (Pober et al., 2001; Kindig et al., 1999; Nathan, 2002). Thus, individual excellence is emphasized, rather than the collaborative efforts required. The organization of AHCs by academic department is designed to facilitate interactions among researchers within the same discipline. Trans- lational research may require the aggregation of expertise across a very diverse set of disciplines, both health and nonhealth related. For example,
AHC AS A TRANSLATOR OF SCIENCE: THE RESEARCH ROLE 87 the evidence-based practice centers funded by the Agency for Healthcare Research and Quality comprise teams with highly diverse expertise and skill sets (Agency for Healthcare Research and Quality, 2002). Organiza- tional boundaries may be crossed not only within the medical school, for example, but also throughout the AHC and even across the university, tapping expertise from economics, engineering, mathematics, psychology, and so on. In one example at an AHC, Dartmouth University created the Center for Evaluative Clinical Sciences in 1989 as a locus for scientists and clinical scholars from across the university who conduct research on issues related to measuring, organizing, and improving the health care system (Center for Evaluative Clinical Sciences, 2003). The center gathers physicians, epidemi- ologists, psychologists, sociologists, economists, medical geographers, stat- isticians, and others to answer such questions as how well medical and surgical procedures actually work, how health care resources are distrib- uted and used, how patients value medical interventions and their conse- quences, and how the quality of medical and surgical care can be continu- ously improved. In another example, the University of Virginia created a Department of Health Evaluation Sciences in 1995 to provide multidisci- plinary scientific and analytical services to its Health Sciences Center and the rest of the university. These services involve examining the development of new approaches and strategies in such areas as the prognosis and clinical and genetic risk assessment of health and disease; medical decision making; and medical practice delivery for individuals and populations (University of Virginia Health System, 2003). Another organizational issue is how oversight of research is performed, which may not match the way research is conducted. Institutional review boards (IRBs) were established in the 1960s (Moses and Martin, 2001). Since then, however, studies have increasingly involved multiple institu- tions and settings. If a study seeks to examine the clinical and cost outcomes of care for a group of patients undergoing a cardiac procedure in a hospital, that study is reviewed by the hospital's IRB. However, if the researchers want to understand outcomes beyond the hospital stay, for example, by examining care at a rehabilitative unit and at the outpatient clinic, the study may encounter separate IRBs for each care setting, even within the same system (Barbara McNeil, personal communication, 2002). The problem is multiplied for multi-institutional studies. These processes add both time and cost to the research effort. IRB processes must be redesigned to ensure adequate protection for study participants in order to gain their trust and participation, without placing an excessive burden on investigators. Since clinical and often health services research involves patients, the pressure to address this issue will grow. Finally, a lack of good information systems is another type of organiza-
88 ACADEMIC HEALTH CENTERS tional obstacle. As discussed previously, translational research in the 21st century will depend on access to good information systems (Manning, 2000). Research in biomedical fields such as genomics generates immense amounts of data to be analyzed. Correlation of genotypes with phenotypes will require access to longitudinal clinical information and large numbers of patients. Additionally, measuring the effectiveness of interventions and as- sessing in both clinical and cost terms their impacts on practice patterns and outcomes often requires overlaying data collection on clinical practice. And enhanced surveillance of disease outbreaks or bioterrorism events will require improved information systems that can link the acute care and public health systems. Such data collection needs to be integrated into care processes and other routine procedures, or the data become too difficult to collect and/or too expensive to harvest. Most studies on effectiveness and outcomes have relied on administrative data for the conduct of retrospec- tive analyses; however, if clinical and health services research is to affect health care policies, practices, and outcomes, more timely data, including real-time data, will be needed (Stryer et al., 2000). To this end, there is a crucial need for better information and communications systems with capa- bilities for knowledge management and decision support. Inadequate Federal Funding of Clinical and Health Services Research Support for research comes from a variety of sources. The bulk of federal funding for health-related research is provided to and by NIH. Funding for NIH grew from $3 billion in 1980 to more than $20 billion in 2001 (U.S. General Accounting Office, 2001), fueling the growth in basic science since, as noted, the majority of funds has been allocated to labora- tory-based biomedical research (Schroeder et al., 1989). In 2001, NIH awarded about $16 billion in extramural research awards, about half of which went to AHCs (National Institutes of Health, 2002a). Health-related research support is also provided by the Veterans Health Administration, the Agency for Healthcare Research and Quality, the Centers for Disease Control and Prevention, the Department of Defense, the Centers for Medi- care and Medicaid Services, the Indian Health Service, the Department of Energy, the Environmental Protection Agency, and even the National Aero- nautics and Space Administration (National Science and Technology Coun- cil, 2000). Estimates of support for basic biomedical, clinical, and health services research vary. In estimating the proportion of NIH support for clinical research, the NIH Director's Panel on Clinical Research used a very broad definition of clinical research that included mechanisms of human disease, therapeutic interventions, clinical trials, development of new technologies, epidemiologic and behavioral studies, and outcomes and health services
AHC AS A TRANSLATOR OF SCIENCE: THE RESEARCH ROLE 89 research (National Institutes of Health, 1997). At that time, the panel estimated that 38 percent of the NIH budget was devoted to clinical re- search. Some believe, however, that this figure overestimates the resources devoted to clinical studies in that some studies may contribute to under- standing of disease without directly involving humans (Schechter, 1998). Other estimates suggest that a lower proportion of NIH funding is devoted to clinical research. For example, an earlier analysis by the Institute of Medicine (1994) revealed that 90 percent of NIH extramural grants sup- ported basic science research and only 10 percent clinical research (Institute of Medicine, 1994). Federal support for clinical and health services research has not been at a level comparable to that devoted to basic biomedical research (Sung et al., 2003). In fiscal year 2000, the total research budget was approximately $21 billion for NIH, the Department of Defense, the Centers for Disease Con- trol and Prevention, the Department of Energy, the Department of Veterans Affairs, the Health Resources and Services Administration, the Agency for Healthcare Research and Quality, the Food and Drug Administration, and the Centers for Medicare and Medicaid Services combined. The budget for clinical research was estimated at approximately $7 billion, plus an addi- tional estimated $1.3 billion dedicated to outcomes and health services research. Funding for health research aimed at populations and commu- nity-based prevention is low, and not a priority in government funding or academia (Institute of Medicine, 2002e). Support for health research is also provided by private industry. Health- related research and development by private industry increased 382 percent between 1985 and 1997 (The Commonwealth Fund Task Force on Aca- demic Health Centers, 1999). Domestic research and development expendi- tures by members of the Pharmaceutical Research and Manufacturers of America (2002) were estimated to total almost $24 billion in 2001. Invest- ment in research by the top 20 pharmaceutical companies has more than doubled in recent years (Morin et al., 2002). AHCs have benefited from this investment by forging partnerships with private industry to conduct com- plex studies and enable each sector to tap the expertise and resources of the other. Health-related research is also conducted by other schools and de- partments of a university, as well as by independent research institutions, consulting firms, managed care plans, hospitals, professional societies, foun- dations, and government agencies (Kindig et al., 1999). It is difficult to obtain precise figures on patterns of overall spending for different types of health-related research. Part of the difficulty is due to definitional overlaps, as noted previously. In addition, however, there is no explicit policy toward shaping research and development in health care (Weisbrod and LaMay, 1999); thus there is no frame of reference, making it difficult to array data consistently across agencies and funders. The devel-
90 ACADEMIC HEALTH CENTERS opment of a well-formulated research agenda for basic, clinical, health services, and prevention research could permit the identification of priori- ties based on good measurement, better public dialogue between policy makers and the users of research on how to incorporate scientific advances into practice and policy, and improved collaboration across the many par- ties that are interested in and benefit from such research (Frist, 2002). Ethical Issues Related to Research New areas of research generate new ethical issues of serious concern. We discuss here only some of the ethical concerns that must be addressed. First, although both AHCs and private industry can bring benefits to research partnerships, issues related to conflicts of interest can arise, includ- ing financial conflicts of interest, potential biases in reporting positive re- sults, researchers' access to complete data, and effects on research priorities (Korn, 2000; Gelijns and Thier, 2002; U.S. General Accounting Office, 2001). Furthermore, it has been suggested that evaluative research requires some distance in the relationship between the developer of a product or device and its evaluator (Gelijns and Thier, 2002). Therefore, there is a potential conflict of interest for AHCs if they are engaged as both the developer and evaluator of new technology. Some AHCs are establishing separate research institutes to facilitate collaborations and reduce conflicts of interest by housing the research outside of the AHC (Moses and Martin, 2001). In general, AHC and industry relationships should be formed with recognition of the potential for conflicts of interest at both the individual and the organizational levels, and with attention to the expectations each party brings to the work and obligations for disclosure (Broder, 2002; Institute of Medicine, 2001, 2002b). Serious ethical concerns also arise from emerging technologies in genet- ics, stem cell research, and cloning, affecting work across the research continuum. Ensuring the confidentiality of patient information while mak- ing it accessible to researchers can be expected to require continuing policy attention. As studies on humans become more complex, concerns also arise regarding how patients can be protected and adequately informed of the risks associated with the research in which they participate (Institute of Medicine, 2001, 2002). The committee also notes the importance of other ethical issues, such as confidentiality of data, conflicts over tissue sampling, and the uniqueness and special handling of genetic data. These issues merit additional study but were beyond the scope of this committee's charge.
AHC AS A TRANSLATOR OF SCIENCE: THE RESEARCH ROLE 91 IMPLICATIONS FOR THE FUTURE Society needs the results of work done across the entire research con- tinuum. Basic research will continue to be important to produce the funda- mental knowledge and tools needed to improve health. However, clinical, health services, and prevention research will also be necessary to improve understanding of how to improve health and to translate the findings of basic research into clinical and community settings so its benefits will reach people. In the coming decades, AHCs will need to increase their emphasis on research across the continuum, including clinical, health services, and pre- vention research in addition to basic research. All such work needs to be recognized, rewarded, and supported. Meeting this need will not require that each AHC expand its research portfolio. One way to encourage greater emphasis across the continuum is through collaboration (Goldstein and Brown, 1997). The individual clinicianresearcher represents one model, but another approach is clinicians and researchers working together. Col- laborations are being created organizationally through large research insti- tutes and centers, but even a few individuals working together can generate important advances. For example, the discovery of the anti-inflammatory properties of cortisone was the product of collaboration among a clinician, a chemist, and a pharmaceutical company (Goldstein and Brown, 1997). As noted earlier, such collaborations, especially across disciplines, are not ad- equately rewarded in AHCs, which emphasize individual achievement. Funding agencies also need to support and foster collaborations across the research continuum. At present, it is difficult for agencies to support interdisciplinary approaches, either by funding research led by an interdis- ciplinary team or by providing interagency funding. Furthermore, the inter- related fields of information technology, biological sciences, and materials sciences may offer some of the most promising research in the future (Frist, 2002) but are not generally linked together by the various funding agencies to maximize their potential. Federal agencies need to improve their coordi- nation in support of both research and research training programs, and to support much-needed collaborations across researchers and institutions.