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 49
The Nation's Physician Workforce: Options for Balancing Supply and Requirements 3 Relationship of Physician Supply to Key Elements of the Health Care System Americans might be said to want sufficient numbers of physicians to enable them and their loved ones to have reasonable access to quality health care services at an affordable cost. Obviously, our nation's record is far from achieving these objectives. The tendency has been to believe that simply adding more and more physicians to the country's present stock will go a long way toward meeting these goals and that doing so would have few, if any, serious negative ramifications. Chapter 2 presents solid evidence that the United States has succeeded dramatically in increasing its supply of physicians—to a level that some, but not all, on the committee characterized as a surplus—and that if an oversupply does not now exist, it will at some future date given current trends in training and utilization. The significance of this phenomenon, in terms of its impact on many other aspects of the nation's health care system, remains controversial, and the issues have been widely argued in the literature and within health policy and professional circles.1 To consider further the impact of a possible mismatch between physician supply and requirements, this chapter examines the nature of the effect of these high numbers on the nation's health care system. The committee examines this
OCR for page 50
The Nation's Physician Workforce: Options for Balancing Supply and Requirements issue from five perspectives: the costs of health care, access to care, quality of care, efficient use of human resources, and the future of academic health centers. The essential questions are the following: Do we know now what effect very high numbers of physicians might have? Will we be able to project the effects with a satisfactory level of precision, especially as the system itself is evolving unpredictably? Do the answers differ depending on what assumptions are made about the degree of penetration of managed care and the size of the residual fee-for-service (FFS) sector? As noted in Chapter 2, forecasts of supply and, especially, requirements are fraught with many uncertainties and difficulties. A small mismatch of supply and requirements in one direction or the other probably should not be taken as evidence of a true or meaningful difference. In this discussion, therefore, the committee considers the above-mentioned issues on the assumption that substantial numbers of physicians are currently available, large numbers are continuing to be produced, and even higher numbers—indeed, a surfeit in most experts' views—can be forecast for the near future. HEALTH CARE EXPENDITURES Using standard models of microeconomic theory, one might expect that a significant increase in the number of physicians would increase competition, produce lower incomes for physicians, and reduce health care costs. Theory, however, is not entirely or consistently borne out in practice. Competition Few data document whether ''competition" in the classic sense—among physicians per se—has increased, although the fact of discounts, sometimes sharp, in physician fees and hospital charges suggests that competitive forces have been at work in recent years. Clearly, the nation's health care system is moving in a rapid and unstructured way to far greater levels of competition than it has ever known. This phenomenon appears mainly to involve competition among health care plans, integrated delivery systems, and hospitals. As the country moves to larger, but fewer, networks and systems and downsizes in certain areas such as hospital beds, physicians will increasingly find themselves attached to or employed by such plans; in principle, competition among them—even if not overt—might increase. Nonetheless, that scenario is somewhat afield of traditional ideas of direct competition among individuals or groups (in the form, for instance, of price wars or limitation of staff privileges). In sum, no clear evidence exists about the nature of the relationship between a
OCR for page 51
The Nation's Physician Workforce: Options for Balancing Supply and Requirements purported oversupply of physicians and competition among them, at least in a managed care environment. Physician Incomes The link between supply of physicians and their incomes (see Chapter 2) is tenuous at best, especially in a health care environment undergoing the many changes in organization and specialty distribution occurring in this nation. No clear evidence exists that an excess of physicians is driving their net incomes down to any significant degree; even if an oversupply did have a dampening effect on incomes, this might not translate into direct, significant dollar savings in the nation's overall health care spending. In any case, the committee has no considered view on the inherent desirability of maintaining current levels of physician income. Health Care Costs It is in the area of aggregate health care costs that theory and practice appear to diverge most dramatically. Especially during the 1970s and 1980s (when FFS reimbursement was more dominant than today), both physician supply and aggregate health care expenditures rose dramatically.2 The very high level of and continuing increases in health care expenditures in this country are matters of significant debate, punctuated by disagreements about the major reasons for the unabated rise in outlays (such as growth in the population, aging of the population, technology advances, general inflation, increasing use of services in the FFS world, and the like). One major factor (as noted in Chapter 2) appears to be that physicians account directly for about 20 percent of health care expenditures in this country and indirectly for 70 to 90 percent of all expenditures—essentially generating outlays every time a patient is seen, a diagnostic test ordered, or a hospitalization advised. As many have noted, under FFS the financial incentive is for physicians and other providers to do more. One can conclude that in a health care world dominated by FFS arrangements and traditional indemnity insurance, a surplus of physicians will be associated with poor controls on costs or expenditures and with overall increases in outlays. As this country moves to managed care, the net (or combined) effect of managed care and a very high level of physician supply is difficult to predict. This is especially so because certain other factors, such as a growing and aging population and ever-improving technology, will continue to have appreciable influence. Moreover, within managed care systems, physicians will remain responsible for most of the decisions that generate the costs involved in delivering care, even if, on balance, managed care can be expected to control
OCR for page 52
The Nation's Physician Workforce: Options for Balancing Supply and Requirements those costs to a much greater extent. By contrast with FFS, however, the incentives in managed care are to do less, including having fewer physicians engaged in patient care. In general, one can conclude that a health care system heavily dominated by a small number of large managed care entities will, all other things equal, not be affected by a physician surplus in the way a FFS-oriented system is. In fact, an oversupply of physicians in a world dominated by managed care catalyzes the cost-constraining effects of managed care, quite apart from the number of services that physicians order or provide. Some argue that the fact of a surplus of physicians has fostered the growth of managed care (and, in so doing, possibly moderated cost increases). Large numbers of physicians permit managed care and integrated delivery systems to engage in efficient business practices. In effect, a surplus enables competition to work on behalf of these plans: for example, managed care organizations can hire (or contract with) new physicians at lower salaries, dismiss or end contracts with older or higher-paid physicians, substitute nonphysician primary care personnel for physicians, and institute significant "gatekeeper" and "referral" controls on the use of specialist physicians. These steps in turn drive costs down, although whether they result in equal or higher quality is not known. Hence, the net impact of a possible physician surplus on health care use and aggregate expenditures will depend in large part on what proportion of the physician workforce continues to function under FFS and what proportion in capitated systems. For example, the proportion of physicians with any managed care contracts rose from 61 percent in 1988 to 75 percent in 1993, but as noted, managed care can cover a wide array of tightly controlled or quite loosely managed plans. Physician impact on utilization and outlays will also depend on other factors such as how much, in the future, FFS plans incorporate elements of utilization management and how much outpatient care is substituted for hospital care (PPRC, 1995). For instance, managed care "norms" may well have a cost-conserving influence on the behaviors of physicians remaining in the FFS sector. In a simple analysis, the committee would conclude that a physician surplus would aggravate costs and expenditures in a FFS-dominated system and would mitigate them in a managed-care-dominated world. Certainly any strong tendency toward induced demand that some see in the FFS system would not likely arise in a managed care world. On balance, however, the committee could not dismiss the contention that a substantial part of the U.S. system will remain oriented to FFS and governed by its incentives for some years to come: managed care approaches may not be viable in some parts of the country or for some population groups, and others will be willing to pay for out-of-plan use themselves. To the extent that this is true, a physician surplus can be expected to be positively, not negatively, correlated with national health care spending.
OCR for page 53
The Nation's Physician Workforce: Options for Balancing Supply and Requirements ACCESS TO CARE Access to care ranks as the second-highest priority issue for health care policy in this country, after expenditures. Considered as "the timely use of personal health services to achieve the best possible health outcomes" (IOM, 1993a, p. 4), access has easily been the longest-running of the three major health issue in this country (i.e., costs, access, and quality). Concerns about access to adequate health care for certain groups (e.g., the elderly, the poor, residents of rural areas) well antedate the dramatic rise in expenditures of the 1970s and thereafter, the appreciable rise in physician supply over the past two decades, and the contemporary focus on quality of care. Two to five decades ago, access problems were seen to be closely related to the perceived undersupply of physicians in the country. Solving the physician supply problem was regarded as a major step in addressing access issues. It appears, however, that the nation "oversolved" its physician supply deficit and yet did little, in the aggregate, to overcome access problems (with the obvious exception of the elderly through the Medicare program) (Politzer et al., 1991). The advent of managed care today muddies analysis of the effect of a physician surplus on access to care, whether the access problem relates to insurance, physician availability in a geographic area, or utilization controls. Some observers foresee that a wholesale movement to tightly managed systems might, in the short run, aggravate an ostensible oversupply problem. Managed care organizations have the mission of tightly controlling access to services, as already noted. Furthermore, health maintenance organizations (HMOs) typically staff at physician-enrollee ratios far below the physician-population ratios now prevalent in the United States. HMOs and other managed care entities vary considerably in their staffing patterns, however. For example, information from the University Hospital Consortium suggests that across five large HMOs, the average number of physicians per 100,000 population was about 138 (in contrast to the figure of 180 cited in Chapter 2 for the nation as a whole).3 Exactly what these figures imply about access to "in-plan" physician services among the populations enrolled in such plans remains to be seen. Overall, however, the express limitation on direct access to physicians (or at least specialists) in managed care entities is not clearly consistent with expanding access to care across all groups in the country. Physician supply and access issues are often examined in terms of two areas of "distribution"—geography and specialty. These topics are discussed only briefly here because the committee was not charged with considering the complex questions of geographic maldistribution or the generalist-specialist imbalance that persists in this country. They do serve to illuminate the complexities of the relationships between access and physician supply, especially when large numbers of people in the country have inadequate access to care and
OCR for page 54
The Nation's Physician Workforce: Options for Balancing Supply and Requirements when the relative balance of FFS and managed care approaches to organization and financing is still uncertain. Geographic Distribution The nation has created a very large physician workforce but has not solved the problem of geographic maldistribution of physicians (or health care services and facilities generally) (Kohler, 1994). This is true for both specialists and generalists. The dramatic increase in the ratio of physicians to population—coming as a result of a national policy that expanded domestic production of physicians and opened the doors wide for international medical graduates (IMGs)—did not ultimately mean that physicians diffused more evenly across the nation or settled in even adequate numbers in places severely short of an appropriate range of physician specialties. More than a decade ago, researchers at the RAND Corporation documented an increased diffusion of specialists into areas having a lower supply (Schwartz et al., 1980; Williams et al., 1981; Newhouse et al., 1982a, 1982b). Despite this movement, few would deny that, today, both rural areas and inner cities continue to face access problems. Indeed, the most recent data from the Council on Graduate Medical Education (COGME, 1995) show that the geographic maldistribution (e.g., for counties of fewer than 50,000 residents) is worsening, not improving. Various efforts have been made over the years to respond to perceived shortages of physicians in rural and poor areas (Desmarais, 1995; Mullan, 1995). For example, state and federal (Title VII) funding of family practice residencies has been considered one useful approach, inasmuch as family practice physicians are more likely than other specialists or generalists to practice in rural areas. This alone, however, is not likely to solve the maldistribution problem for rural populations. Another such program is the National Health Service Corps, which places physicians in so-called Health Professional Shortage Areas (HPSAs).4 HPSAs are identified on the basis of several factors: low physician-population ratios; high rates of poverty, infant mortality, and low birthweight; and poor access as reflected in waiting times, visit rates, or distances to care. As of 1994, nearly 2,740 HPSAs had been designated, of which about 67 percent are rural; this number is more than double that for 1978 (1,242 HPSAs, of which 72 percent were rural) (BPHC, 1995). Altogether, today they represent a need for more than 5,340 primary medical care personnel for a population of not quite 48 million individuals.5 In principle, a large surplus ought to enrich the pool of physicians from which National Health Service Corps and similar programs can draw (at least if fewer attractive loan programs were available). The constraints on solving some access problems in this way, however, lie in the number of such
OCR for page 55
The Nation's Physician Workforce: Options for Balancing Supply and Requirements positions that can be supported through federal appropriations for those programs and in the structure of the financial incentives, not the size of the physician pool. Some observers have conjectured that greater numbers of foreign-trained physicians might help alleviate some of the geographic maldistribution problems in this country. That, too, has not been borne out. By and large, IMGs go to nonmetropolitan areas only as frequently as domestic doctors (Mullan et al., 1995). No solid data indicate whether IMGs are or are not more likely than their U.S. medical graduate (USMG) counterparts to be found in the nation's inner cities. The consensus view appears to be that the geographic distribution of IMGs parallels that of USMGs or is, if anything, slightly skewed toward the largest metropolitan counties in the nation (Mullan et al., 1995).6 In short, the presence of more and more physicians in this country has not solved problems of access to care. Although a greater supply of physicians has helped improve access to specialists in smaller communities, it has not completely solved the problem or corrected perceived imbalances in supply across geographic regions. However, access difficulties probably never were, and never will be, wholly amenable to resolution through increased "supply" when significant financial and nonfinancial barriers to seeking care continue to exist. Furthermore, if the nation's health care system becomes heavily dominated by managed care, arguably some access problems may well be exacerbated if such organizations do not find it economically attractive to establish networks in presently underserved, frontier, or similar areas. Of course, eventually a huge oversupply of physicians might well induce movement into underserved areas; for instance, anecdotal evidence is coming in of physicians' leaving California to practice in Wyoming, Montana, and similar rural states (where, arguably, FFS medicine still predominates). In the committee's view, however, attempting to solve the access problem simply by creating an oversupply sufficient to induce such behavior is an extremely inefficient way to proceed. The nation requires specific strategies and policies to address its continuing access problems that do not depend on the numbers of physicians it has. Specialty Distribution Although experts have recognized for many years that the country has a shortage of generalist physicians, most experts now believe that it is relatively modest in size, and some even hold that the generalist supply is in balance with or slightly in excess of requirements. Certainly when other types of health care professionals (e.g., advanced practice nurses, nurse practitioners [NPs], physician assistants [PAs]) engaged in primary care are added to the equation, most authorities would take the position that no shortage exists in personnel capable
OCR for page 56
The Nation's Physician Workforce: Options for Balancing Supply and Requirements of delivering primary care services, although they may still disagree about whether the numbers are at present at, or slightly above, desirable levels.7 There does seem to be widespread agreement about an oversupply in many of the nonprimary care specialties and subspecialties. To the extent that too few physicians today are in primary care and too many in specialty and subspecialty practice, this specialty maldistribution will persist for some time (COGME, 1994, 1995). Whether these imbalances will be exacerbated or ameliorated by broader developments in the health care arena is a matter for some debate. If data from some mature managed care markets are any indication (see Chapter 2), the oversupply of specialists is quite apparent, simply because managed care organizations use many fewer specialists than are practicing in the present FFS world (Weiner, 1993; Bruce Sams, Belvedere, California, personal communication, 1995). Such information suggests that an aggregate oversupply of physicians will not solve the problem of maldistribution by specialty and that the problem of too many specialists and too few generalists might worsen if managed care entities proliferate. As with geographic maldistribution, the infusion of large numbers of IMGs did not, and in the future will not, help resolve specialty imbalances because IMGs subspecialize as frequently as USMGs; only about one-third of both USMGs and IMGs in patient care elect primary care specialties of general or family practice, pediatrics, and internal medicine (Mullan et al., 1995). The complicating factor is that the "pipeline" of physicians in these fields—that is, the numbers entering into or presently in training—is quite long. These patterns imply that an oversupply of physicians, particularly of specialists, will lead to large numbers of IMGs' and domestically trained doctors' competing for positions in urban areas in preference to rural and poor communities and for sufficient work to maintain their skills. The committee did not foresee large-scale movement of either IMG or USMG specialists into underserved areas as a means of meeting either the primary care or the specialty care needs of those populations, given the lack of a viable FFS infrastructure (e.g., appropriate hospitals) in many of those locales and the likely low use of specialists by any managed care organizations that might enter those areas. Furthermore, the factors that induce or ameliorate specialty maldistribution may be more complex than those affecting geographic maldistribution. Among the changes that might affect specialty maldistribution are choices that physicians already in practice (or about to enter practice) make about what they call themselves and what they do. For example, some physicians with mixed practices (e.g., those certified in both internal medicine and one of its subspecialties) may increase their care of general problems; other specialists may seek to retrain in a generalist discipline via short courses and thereafter market themselves as generalists. Whether these practitioners will then be as expert or
OCR for page 57
The Nation's Physician Workforce: Options for Balancing Supply and Requirements interested in their subspecialty practices as their colleagues who devote full time to their subspecialty practice is a matter of conjecture. One additional point underscores this uncertainty: generalists who have completed full residencies in generalist programs are more likely to disparage "retraining" as a poor substitute for their own training. Finally, many managed care programs are explicitly built to distinguish primary care gatekeepers from subspecialists and to keep them clearly separated. Their formal generalist-specialist policies may be very important in determining how these changes eventually play out. In short, the nation has a serious disproportion of specialist to generalist practitioners even with its very large numbers of physicians. Increasing the supply even further (which some would characterize as perpetuating a surplus) will not likely correct this imbalance. A continued mismatch of specialty supply to requirements, however, especially if stringent managed care operations begin to dominate U.S. health care, may risk a number of unforeseen consequences.8 Furthermore, the workforce issues for specialty and primary care medicine differ. Global solutions that ignore the primary-subspecialty issues should be avoided, so that they do not do harm to the developing primary care base of our health care system. Summary Comment About Access to Care What seems clear is that (1) the present large supply of physicians has not resolved long-standing problems of geographic or specialty maldistribution; and (2) ever-greater numbers of physicians per se has not yet alleviated the problem of access to basic, decent health care for growing numbers of people in this country.9 Eisenberg (1994) sums up the issue as succinctly as any: "trickle-down" approaches to solving access problems through increased supply have not worked. As managed care approaches to the organization and financing of health care diffuse across the country, the possibility of an even greater surplus of physicians cannot be dismissed, and the existence of many more physicians than appear to be required may do little to redress serious problems of access to care in the nation as a whole. QUALITY OF CARE According to a 1990 Institute of Medicine (IOM) report, quality of care is "the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge" (IOM, 1990, p. 21). That report goes on to document ways in which health care quality may be called into question: (1) use of unnecessary or
OCR for page 58
The Nation's Physician Workforce: Options for Balancing Supply and Requirements inappropriate care; (2) underuse of needed, effective, and appropriate care; and (3) poor performance in the technical or interpersonal aspects of care. As can readily be appreciated, an oversupply of physicians (either specialists, generalists, or both) would have different implications for different elements of questionable quality of care. Few reliable data are available to clarify what the net effects might be of the numbers of physicians the nation now has. The picture becomes murkier when the rapidly changing environment of physician practice—especially the accelerating growth of managed care systems—is taken into consideration. High Numbers of Physicians Overall General Effects Data to answer the question of whether large numbers of physicians will have a beneficial, a harmful, or no net effect on quality of care are exceedingly scarce. As implied in the discussion of expenditures, a significant physician oversupply would pose potential harms to patients from unnecessary care (Perrin and Valvona, 1986), especially in the FFS sector. These harms can range from simple inconvenience and the financial burdens of uncalled-for out-of-pocket costs to severe iatrogenic complications. Conversely, excess numbers of physicians might, in accordance with economic theory, drive out poorer performers, leaving only very good physicians who could be expected to render high-quality care. The interactions of physician supply with a FFS or a managed care orientation to health care delivery, in terms of quality of care, are complex. Evidence suggests that the quality of care in traditional HMOs is in general equivalent to that in nonmanaged care (e.g., FFS) arrangements, at least for average populations, although this may not hold for persons who are both poor and sick (Newhouse and the Insurance Experiment Group, 1993). According to Safran et al. (1994), results from the Medical Outcomes Study on core dimensions of primary care (IOM, 1978) indicated that HMO patients had better financial access and coordination, but poorer organizational access, continuity, comprehensiveness, and judgments about provider accountability (technical skill and interpersonal manner). These studies, however, are based more on traditional types of prepaid, capitated systems, not on the emerging hybrid or for-profit models. The existence of large numbers of practitioners might well mean that managed care organizations could take advantage of physician competition and hire or contract with only the "best" physicians (e.g., those who are well trained and experienced); equally plausible is that they might employ the youngest, least experienced, or cheapest physicians, who might or might not be the best. To the
OCR for page 59
The Nation's Physician Workforce: Options for Balancing Supply and Requirements extent that tightly managed systems (or plans with very high out-of-plan cost sharing) succeed in discouraging enrollees from seeking care from physicians not in the network or plan, the possibility that persons would have to settle for potentially poorer care from network physicians cannot be dismissed out of hand. Volume and Quality Higher volume of services rendered (e.g., operative procedures) has long been regarded as directly related to higher quality of care (OTA, 1988, Chapter 8). Data are especially persuasive about open heart surgery. No definitive volume requirements or norms for different types of surgery or other health care services have ever been promulgated, however. Poorer surgical outcomes are more likely to occur where surgeons have small workloads that preclude their performing the volume of procedures necessary to maintain efficiency and technical skills (Luft et al., 1990). Problems of quality can arise if physicians practice outside their usual areas of competence, for whatever reasons (as noted in the discussion about specialty distribution), or if their skills erode with little practice. Physicians who are "finding things to do" to compensate for being less than fully occupied may cause iatrogenic illnesses as well as waste scarce resources. The volume-quality relationship is tempered to the extent that unnecessary and inappropriate interventions are given—for instance, to persons who could be equally well served by "watchful waiting." Consequently, in the FFS system, too many physicians may indeed be associated with higher volumes of care but not necessarily with better patient outcomes or overall higher levels of quality. This particular side effect of FFS incentives might not operate in a managed care environment, where a different picture emerges for managed care plans with tight controls on utilization of medical or surgical services. In these circumstances, volumes may well decline for individual practitioners, with possible deleterious consequences for quality. In general, the committee agrees with the notion that higher volume is related to higher quality of care. However, because of the uncertain balance between managed care and FFS arrangements in coming years—and the very different effects those approaches have on volume of services—the net impact of a physician surplus on volume of services, and hence on quality of care, cannot be foretold with any certainty. Provider and Consumer Satisfaction Another, less-well-appreciated potential effect involves physician and public attitudes, but the net impact of large numbers of physicians is not simple to
OCR for page 64
The Nation's Physician Workforce: Options for Balancing Supply and Requirements confidence—among them the extent of managed care and how tightly such systems control access to physicians, the choices that doctors in training as well as those in practice make about primary or specialty practice, and the underlying proportion of physicians in the country who are schooled in foreign institutions. In the end, it seemed as if overall levels of quality of care will be driven by traditional professional attitudes, consumer expectations, and effective quality improvement programs more than by sheer numbers of practitioners today or tomorrow, but the committee nonetheless notes that the pipeline for producing physicians is a very long one, so lack of action today may have real and undesirable consequences for tomorrow. OTHER ISSUES IN PRODUCTION AND UTILIZATION OF THE PHYSICIAN WORKFORCE Human Resources Considerations Opportunities for Medical Careers Physicians in the United States come from what are arguably the nation's brightest young adults. A major concern is that these aspiring and talented youth will invest time, energy, and financial resources to become physicians, only to find themselves incapable of earning a living or practicing to the standards to which they were educated. This is a clear waste of talent and human capital investment. Prospective medical students are not turning away from such careers on the basis of information available to them today. For example, applications to medical schools are at an historic high; for the 1994-1995 academic year the ratio of applicants to accepted students was 2.6 to 1 (Barzansky et al., 1995), up from 2.5 the year before (Jonas et al., 1994) and 2.1 and 1.9 for the two years before that. Not since the 1974-1975 applicant class has the applicant acceptance ratio approached the current level (Barzansky et al., 1995). The policy question is whether the investments that they, their families, and the nation will make in such education will pay off for them and for society as a whole. Unlike a graduate degree in law or business administration, medical education is not of great use as a "foundational" degree. Doctorates in medicine are nearly uniquely helpful to those who intend to dedicate their lives to the healing arts. That is not to say that some fields, such as literature, have not benefited from the insights of individuals who come to them with medical training, and the occasional individual may deliberately shift his or her medical career into another pathway. Apart from some physicians who engage in basic or applied scientific or health services research, however, full training through
OCR for page 65
The Nation's Physician Workforce: Options for Balancing Supply and Requirements a residency may well be unnecessary, inappropriate, and excessively costly in human and financial terms if the individual does not then pursue a medical career. In addition, over the longer run, the possibility exists that very talented young Americans will decide not to pursue a career in medicine. This could be especially true if information and market signals suggest that they may be underutilized, be unable to practice in the locale or specialty of their choice, or have to compete with IMGs who might well accept lower stipends, fees, or salaries simply to stay in the United States. Although market signals should indicate clearly the prospects of potential students, one would not wish them to be so pessimistic as to turn away the best candidates. However, whether alternatives to medicine will be attractive career choices for such students is uncertain and likely to vary over time as economic and other factors in different professions and sectors vary. The issue involving IMGs presents a particularly knotty dilemma as their numbers in this country rise (quite irrespective of quality-of-care differences). The two major concerns are as follows. First, IMGs will continue to come to the United States as full-fledged physicians or for training and then remain to practice—especially in the numbers experienced to date. Second, through these mechanisms, IMGs will contribute to a real oversupply of physicians and, indirectly, to the market signals to prospective USMGs. These factors could have the unintended effect of denying the youth of this country the opportunity for a medical career.13 The interplay between undergraduate and graduate medical education complicates the picture. One option sometimes suggested for trimming the output of physicians in this country is to reduce the number of individuals in undergraduate training, either by closing medical schools or by reducing class sizes. However, the demand for graduates to fill residency slots remains high, and as noted in Chapter 2, the excess demand is filled with IMGs—currently about 7,000 or more residency positions a year because the production of M.D.s and D.O.s (about 17,500 per year) falls that short of the number of residency slots. It is most unlikely that allopathic and osteopathic schools in the United States could be persuaded unilaterally to shut down or dramatically to reduce their class sizes when such an enormous call for their graduates exists. Furthermore, if U.S. medical schools closed or reduced their class sizes, the net result in the current environment would simply be to increase the numbers of IMGs in residency training—arguably, thereby, substituting IMGs who have, on average, less adequate undergraduate or graduate training for generally better-trained USMGs in the overall pool of physicians in this country. As discussed in Chapter 4, the committee would not subscribe to such a solution to a potential oversupply of physicians in this country. In sum, opportunities to practice medicine will be limited if current trends in physician supply (and managed care) continue. To the extent that this is true,
OCR for page 66
The Nation's Physician Workforce: Options for Balancing Supply and Requirements the potential displacement of U.S.-educated physicians by foreign-trained physicians is regarded with dismay by this committee.14 As noted in Chapter 1, the committee adopts the principle that in the constrained professional environment that may be coming, opportunities for advanced medical training ought to be reserved to graduates of U.S. schools (virtually all of whom will be U.S.-born individuals), even though the committee recognizes the legitimate role of graduate training for those IMGs who return to their own countries of origin. U.S. Policy and the World Output of Physicians The world is producing more physicians than the United States can or would be willing to assimilate. Many graduates in other countries, like their U.S. counterparts, are not motivated to practice in areas of medical need in their own nations (Yang and Huh, 1989). Thus, as in the United States, production of physicians in other nations is leading to oversupply there, with various effects. One notable consequence is a significant "brain drain," as trained or prospective physicians from abroad come to the United States in numbers equivalent to the entire graduating classes of 50 medical schools around the world. Although some nations are establishing a balance between production of physicians and expected requirements, most are not. The list of countries that are setting or have already set restrictions on the number of physicians they are educating and allowing to specialize includes Belgium, Canada, France, Germany, the Netherlands, and Spain (Schroeder, 1984; Foley, 1994), largely owing to the conviction of some experts that nothing is more central to cost control of health care expenditures than the aggregate supply of physicians (Schroeder, 1984). Most countries, however, are not taking such steps, and these include the donor nations for most of the IMGs coming to the United States. A U.S. policy with respect to IMGs as open as the current policy provides little reason for those nations to bring their educational output into better balance. Future of Academic Health Centers General Issues Threats to academic health centers. In its July 1995 meeting, the committee discussed at some length the concerns of several of its members about the future of academic health centers (AHCs) in this country. (For in-depth discussions of these issues, see Iglehart, 1994, 1995a, 1995b; UHC, 1995.) Such academic institutions, numbering more than 100, are a valuable national resource. AHCs carry out the bulk of academically based biomedical, clinical, and health services
OCR for page 67
The Nation's Physician Workforce: Options for Balancing Supply and Requirements research in the United States; train essentially all types of health professionals and midlevel and technical personnel at the undergraduate or graduate level; and deliver most of the very high-technology emergency and tertiary care available here as well as considerable amounts of primary care for certain populations. AHCs provide approximately 50 percent of the revenue supporting medical education in the United States through their clinical practice plans, and they provide about 44 percent of all charity care in this nation. In short, they are a linchpin for the nation's health care system, and they provide a high percentage of the world's collective scholarly effort in biomedical and health technology research and development. Some AHCs are under extreme financial stress for several reasons. Their unique missions add to their intrinsic costs, and decreasing revenues from the delivery of medical services (e.g, by the faculties of such academic institutions) will no longer help to underwrite education and research programs (Iglehart, 1994; Van Etten, 1995). Moreover, they tend to have disproportionately more indigent patients and more complex illnesses to treat than most other hospitals. In a managed care world, they become particularly ill-suited to compete in any bottom-dollar way with other institutions without these characteristics, especially in competition with for-profit managed care plans when the "medical loss ratio" (the cost of medical services as a percentage of the revenue from premiums) is an important market criterion for purchasers of health care (Iglehart, 1994). Finally, proposals to lower or reduce indirect or direct medical education payments (or both) add another fiscal pressure. Other developments in the health care sector aggravate the problems that AHCs face. For example, some members of the committee see an increasing antagonism of private practice physicians toward AHCs—sufficiently so that referrals to such centers are dropping. Should such attitudes prevail, decreasing revenues from such referrals would further undermine the support for education (of all health professionals, not just physicians). Others believe that the country already has an oversupply of physicians and that it will worsen, meaning that public policies will tend to favor decreased production of physicians at the undergraduate level (let alone the graduate level). It is not clear how such downsizing might take place, but some see a rising pressure to close (or at least merge) some centers. The effects on AHCs of FFS or managed care approaches to organizing and financing health care are very complex. A significant FFS presence might foster the current configuration of AHCs, whereas a predominant managed care orientation might well undermine it, especially to the extent that managed care enterprises opt to create their own residency or graduate training programs to suit their particular norms and ways of organizing health care delivery. Although the committee did not have time to explore these issues in any depth, it did choose to go on record as concerned about this vital element of the nation's health care system.15 The challenge is to balance the size and
OCR for page 68
The Nation's Physician Workforce: Options for Balancing Supply and Requirements investment in such centers with the nation's (and indeed the world's) requirements for research, education, and health services. This committee would be especially concerned if uninformed debate about a physician oversupply were to feed the notion that AHCs are dispensable. Continued support for AHCs must rest on more than their role in the development of the health care workforce—in particular, their vast contribution to biomedical and clinical research must be understood and sustained. Financial support for education. One important factor in physician education is its public financial support. No other group in the workforce receives direct financial subsidies of the size and scope of those directed at physicians. For example, the federal GME subsidy in fiscal year 1994 was nearly $6 billion (Pew Health Professions Commission, 1995a). Good public policy reasons exist for this assistance. As a case in point, the education of physicians is the longest, most intensive, and most expensive of any educational program of professionals. Moreover, the costs of education are such that the nation would not have a physician workforce even roughly representative of the general citizenry without such subsidies. That is, only the children of the wealthy could reasonably aspire to become physicians without the substantial levels of public and/or private financial support now available. Federal support for GME also provides compensation to AHCs with respect to the severity of illness of their patients and the scope of services they render. Even if, however, one adopts the view that helping to defray some of the costs of physician education is prudent, one can still ask whether the nation is spending more, in aggregate terms, than necessary to meet national requirements for physicians. This question is particularly germane for two reasons. First, the nation is training more physicians than it needs. Second, national budget deficits, which are driven in large measure by health care spending in the public sector, are at unsustainably high levels and are expected to continue to rise. Thus, many observers believe that in terms of national fiscal responsibility, investing in the production of physicians beyond levels needed to meet prudent estimates of requirements is inappropriate, if not foolish, because it wastes both personal and public human capital investments. Whether market-driven or regulatory steps (or some combination) are called for to achieve a good balance is a matter for further debate (see Chapter 4). The Service-Training Link and IMG-Dependent Institutions An additional complexity is the long-standing federal policy that connects payments for graduate medical education—that is, residency training—to service and patient care. These payments, through Medicare reimbursements for direct
OCR for page 69
The Nation's Physician Workforce: Options for Balancing Supply and Requirements medical education and indirect medical education, provide ample incentive for hospitals (including AHCs) to keep the number of residency slots high and, when necessary, to fill them with IMGs. Some in the health field argue that this level of service from residents, whether U.S. or foreign trained, is essential to the provision of inpatient care in underserved areas or for vulnerable populations such as those in inner cities or rural communities. Some also argue that without IMGs in these residency programs, the hospitals could not provide even basic levels of care to their patients. The vast majority of the nation's hospitals, however, are not dependent on IMG residents in this way. Whitcomb and Miller (1995) provided data suggesting that of the nearly 6,000 hospitals in this country, only 77 can be regarded as (a) having substantial residency programs (especially in primary care and major specialties); (b) being heavily involved in providing care to the poor (i.e., among the so-called disproportionate share hospitals); and (c) being IMG dependent.16 To the extent that solving the broader issues of physician oversupply involve direct actions regarding the influx of IMGs to this country, special or targeted efforts can be made to help this small number of hospitals find other solutions to their difficulties of staffing and service delivery. The committee returns to this point in Chapter 4. CONCLUSION If the nation had to choose between too many physicians and too few, it would prefer an excess to a dearth. Obviously, the profession of medicine wants to have the nation adequately supplied with physicians, but it sees little to be gained from a huge imbalance between supply and requirements, especially if circumstances adverse to cost, quality, or access were to result. The committee agrees with this assessment—recognizing that an accurate balance between physician supply and societal requirements is an unachievable goal and generally favoring too many rather than too few physicians. The committee has concluded, however, that the weight of the evidence presented in this chapter and Chapter 2 supports the position that the present oversupply of physicians in this country poses, if anything, more problems than solutions to the nation's health care issues. Among the points to be emphasized are the following: No firm evidence can be marshaled to show a beneficial effect of a physician oversupply on costs, access, or quality. Specifically, it is difficult to see that an oversupply will have much effect on problems of access to care in this country, and an abundance of physicians will not solve the problems of maldistribution by geographic area or specialty. Furthermore, the evidence is mixed about the impact of a substantial oversupply on either
OCR for page 70
The Nation's Physician Workforce: Options for Balancing Supply and Requirements the quality or the costs of care, in part because the effects of health care restructuring toward much greater penetration of managed care are unpredictable. Generally, it is not possible to demonstrate that too many physicians will improve the quality of patient care; indeed, if the surplus is made up largely of IMGs, it may dilute quality. In addition, a surplus will contribute to higher aggregate health care costs at least as long as the nation has a significant FFS sector. Having far more physicians than is needed to meet the nation's requirements is a waste of the federal resources currently spent on physician graduate education, and it may also result in a poor personal investment on the part of prospective medical students. When individuals pursue a medical career in the face of a significant oversupply of physicians, their underemployment or underutilization is a tremendous waste of human resources for the nation. Use of large numbers of IMGs here deprives other nations of their own talent and decreases opportunities for able young persons from the United States to enter the medical profession. Thus, the committee believes that however a better balance is to be achieved, it is in the national interest to avoid a serious oversupply of physicians. The options for doing so are taken up in the next chapter. NOTES 1. The committee opted, for this short statement, not to develop a lengthy critical review. Relevant, albeit selected, articles on various aspects of these issues for the past 15 years or so (i.e., dating to the landmark GMENAC report), other than those already cited in this report, include Ginzberg et al., 1981; Tarlov, 1983; Harris, 1986; Ginzberg, 1989, 1992; Schwartz et al., 1989; Dranove and White, 1994; Moore, 1994; Epstein, 1995. 2. Evidence supporting this link comes from Canada (Barer et al., 1989). Recently, Canada has moved to a rigidly controlled system of aggregate supply as an explicit strategy to reduce the growth of health system expenditures. This approach involves monitoring the numbers of medical school and graduate training slots, and it proscribes international medical graduates from settling in that nation. These policies are quite recent, so it remains to be seen how effective they will be, but clearly Canada has concluded that only by intensive regulation of physician supply can it achieve national health policy objectives with respect to costs. In view of the fact that Canada has little managed care and is basically a FFS system, however, it may offer less of a direct lesson for the United States today than might have been true a decade or so ago. 3. Unpublished information being compiled for the IOM study on the future of primary care also reflects great variation across managed care organizations; it suggests that some large HMOs may be staffing at levels not that far removed from the national
OCR for page 71
The Nation's Physician Workforce: Options for Balancing Supply and Requirements national average (i.e., 180 per 100,000 population), although still below the national average for metropolitan areas (i.e., about 200 per 100,000). 4. Health Professional Shortage Areas (HPSAs) were formerly known as Health Manpower Shortage Areas until passage of the National Health Service Corps Revitalization Amendments in 1990; they superseded Critical Health Manpower Shortage Areas that had been in effect since 1972. HPSAs are divided into the types (primary medical, dental, and psychiatric or mental health) depending on the type of shortage. Similar approaches to designating federal shortage and underservice areas are used by different programs and include Medically Underserved Areas/Populations, Nurse Shortage Areas, and High Migrant Impact Areas. For a history of federal approaches to shortage area designation and criteria, see Lee (1991). 5. In some urban areas, large academic health centers operate community health centers, school clinics, and the like through various types of ambulatory care networks aimed at the poor sections of metropolitan areas. One such system is operated by the Montefiore Medical Center, whose president reports that it is becoming increasingly easier to attract young U.S. medical graduates to the network for reasons that include good practice sites, competitive salaries, hospital privileges, and the growing scarcity of professional opportunities in more affluent communities. 6. Mullan et al. (1995) note that hospital-based IMGs are slightly more likely than USMGs to be in "noncore counties of metropolitan areas … and in select categories of nonmetropolitan counties," presumably in residency and staff positions in smaller communities; the authors speculate that when IMGs are "free" to establish office practices upon completion of graduate training, they opt for the same urban settings as their U.S.-schooled counterparts (p. 1525). 7. Observers also disagree about the net effects of managed care on the use of primary care physicians or nonphysician personnel, noting that shortages of the latter or convergence in incomes may play a considerable role in how managed care plans and integrated health delivery systems configure their primary care teams. 8. Experiences in Canada, Mexico, Germany, and Israel may also give some insight into what could develop in this nation as a result of substantial subspecialist oversupply (Barer et al., 1989). For example, the United States may well witness a sharp division between hospital-based and ambulatory-setting-based physicians, as is the pattern in most other postindustrial economies. In this country, specialists are not without power, and they may be able to move generalists from the hospital setting by reducing and eventually closing off their hospital privileges. In other nations, this cleavage within the medical profession has served to increase retesting and create a hierarchy of status that works against integration and continuity of patient care (Frenk et al., 1991). Some reports from other countries also suggest that with an overabundance of specialists, special prices begin to appear that may give certain patients access to more—and better—attention within the health care system. 9. Access to and demand for health care services may become more problematic in the near future because of the rapidly growing numbers of people in the United States who are now uninsured, underinsured, or losing their health insurance. According to Biles (1995), today upward of 40 million people in this country lack any health insurance. When those who have only sporadic coverage are included, then more than 60 million persons may be uninsured at any one time. Furthermore, approximately
OCR for page 72
The Nation's Physician Workforce: Options for Balancing Supply and Requirements 100,000 persons a month are losing coverage. Certainly a rising number of physicians will not be a solution to these structural problems of shrinking insurance coverage, and the probable lack of effective demand for services might well exacerbate the perception that the nation has a significant oversupply of physicians. 10. Every state now requires at least one year of postgraduate training for licensure purposes. Many, although not all, physicians in the United States seek training sufficient to enable them to become certified by a specialty board in a particular field (including both generalist fields such as internal medicine or family practice and very high technology subspecialty fields). Intuitively, board certification suggests that such physicians would provide higher-quality care. According to one committee member, 85 percent of the practitioners working for managed care organizations are board certified, as contrasted with 62 percent of the general physician community (George Sheldon, University of North Carolina, personal communication, September 12, 1995). In 1988, however, an OTA review of this topic concluded that such certification will not reliably predict which physicians will render high-quality care, and which will not, within a given specialty, and the simple fact of being board certified would likely be an even less reliable indicator of quality across specialties. 11. Evidence from the site visits and public hearing held by IOM committee for the large-scale study on primary care (which is exploring training and education for generalists in some depth) also supports this contention; its report will be available in early 1996. In addition, a different IOM committee is studying the adequacy of nurse staffing (both numbers and skill mix) in hospitals and nursing homes; here, too, preparation of nurses for providing services in both institutional and ambulatory settings is an issue (IOM, 1996, forthcoming). 12. Physicians taking Step 3 of the USMLE have successfully passed Steps 1 and 2. The percentages of USMGs and IMGs successfully passing Step 1 in 1994, on their initial try, were 91 and 51 percent, respectively. For those taking Step 2 examinations in 1993-1994 (which are given at three different times), percentages were—again for those attempting the test for the first time—92 and 47, respectively. For more detail, see NBME (1995) and equivalent issues of this bulletin in past years. 13. A more complicated argument about the deterrent effect of large numbers of IMGs on the U.S. physician supply was raised in committee discussions. It stemmed from two propositions: (1) that foreign-trained physicians who practice here may earn salaries far in excess of what they could command overseas, and (2) that this "subsidy" to the incomes of foreign graduates, which is not effectively available to U.S. graduates, might drive down existing or prospective salaries of USMGs. This phenomenon in turn could send a quite unintended market signal to prospective U.S. medical students. 14. The committee debated with some vigor two additional questions: (1) whether the essential characteristics of an "American" physician are forged in residency training or in college and undergraduate medical education (if not earlier), and (2) whether examinations and similar steps fully capture the range of dimensions and commitment to this nation's social ethic that Americans regard as important in their physicians. If the answer to such matters is that the entire educational process in the United States is unique (with respect to producing physicians) and is something that Americans value, then this may be a sufficient rationale for being concerned about the influx of IMGs.
OCR for page 73
The Nation's Physician Workforce: Options for Balancing Supply and Requirements If, however, such a rationale cannot be sustained, then the issue of IMGs in the physician workforce of this nation should perhaps be examined on international trade policy grounds. Many people in this country subscribe to some version of a free-trade invisible market hand) that services, commodities, and the people who produce them are as inexpensive, convenient, and of high quality as possible. This would imply that the (relatively expensive) production and use of USMGs could rationally be reduced in favor of the (relatively inexpensive) importation and use of IMGs. because this economic decision would (in theory) benefit consumers of health care most. This argument has three weaknesses: (1) it places a great deal of weight on benefits to consumers and less on the interests of other parties who may have different perspectives on the matter; (2) clearly flies in the face of the political reality that Americans value the opportunities for their children to enter pretigious professions. In the end, the committee had neither the time nor the resources to explore these issues thoroughly. It returned to a position of what good policy for the United States, based on the twin arguments that medical careers ought to be appropriately available to persons graduating from U.S. schools and that the education provided in those schools does convey objective and subjective experiences that matter in this country and are not likely to be conveyed in foreign schools. 15. In 1994, the Institute of Medicine organized a Roundtable on Academic Health Centers, which explored these issues in considerable detail over several meetings. A summary of that effort is due in late 1995. One major concern of roundtable members was that the "playing field" for AHCs had to be leveled. That is, such centers cannot easily demonstrate that they are cost-competitive (controlling for quality of care as well as their research and education missions); thus, they would not be able to compete successfully with other health care systems and managed care entities in the community. Some innovative arrangements are emerging that would integrate AHCs into health plans in ways that link them to community-based providers, but gathering further information on these was beyond this committee's resources and charge. 16. The Whitcomb and Miller (1995) analysis was confined to hospitals with residency programs in six specialties and to those that are potentially dependent on resident IMG physicians to provide basic medical care services to the poor. Under these restrictions, the analysis reduced to 688 of the nation's hospitals; 77 of these met all of the authors' conditions for being IMG dependent and providing care to a significant proportion of poor patients. These hospitals are mainly in the nation's metropolitan areas (26 in New York State alone) and serve some of the most vulnerable people in this country.
OCR for page 74
The Nation's Physician Workforce: Options for Balancing Supply and Requirements This page in the original is blank.
Representative terms from entire chapter: