| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1995;60:1541-1546
© 1995 The Society of Thoracic Surgeons
Health Policy Institute, Medical College of Wisconsin, Milwaukee, Wisconsin
Abstract
Medicine is entering an unprecedented era of provider abundance, including both physician and nonphysician providers. Over the next several decades, the projected number of primary care physicians will be more than adequate to meet national needs, although there is no assurance that any number of physicians will create an equitable distribution. At the same time, a growing surplus of specialists is projected. A balanced abundance in both primary care and specialty medicine will continue if approximately 33% of first-year residents ultimately practice primary care and 67% become specialists. In contrast, a shift to 50:50, as has been proposed by the Committee on Graduate Medical Education and others, will lead to a superabundance in primary care and a potential deficiency in specialty medicine later in the 21st century. Under either scenario, maintaining balance will be aided by those physicians with sufficient generalist skills to enable them to practice at the interface of primary care and specialty medicine, the domain of ``middle care.'' The nation will be well served by educational policy that imparts such generalist expertise to medical students and that creates a workforce of highly skilled physicians capable of caring for patients in the technologically advanced clinical environment of the future.
This article focuses on the size and composition of the physician workforce and on the issues facing us as we attempt to modify our training programs to meet future physician workforce needs. Many of these comments are drawn from two recent articles on this subject [1, 2].
The starting point is the general view that better balance is needed in the physician workforce, particularly in the number of primary care physicians relative to the number of specialists. I share that view. Unfortunately, that view has been translated into a specific call for federal regulation that would mandate that by the end of this century half of all United States medical graduates would pursue careers in primary care: the ``50% solution'' [3]. Moreover, the imperative to do something is strong.
Many factors underlie this regulatory approach. Some are philosophic, others economic, and some are political. Both cost and access are prominent in the discussion. I agree there must be an increased emphasis on primary care and have worked to strengthen primary care during my 9 years as dean at the Medical College of Wisconsin. I am also sympathetic to the need to control healthcare costs and improve access to care. However, I do not think the 50% solution is a good idea, and I am opposed to regulation of the physician workforce [2].
At the outset, I want to emphasize three points. The issues that have fueled the physician workforce debate are complex, and only some of them are amenable to solution through changes in the workforce. The data underlying workforce analyses have significant limitations. Finally, physician workforce projections are laced with uncertainty. Complex issues, limited data, and uncertainty—these confound the process of formulating physician workforce policy.
Healthcare Providers
The healthcare workforce is not merely composed of stereotypic primary care physicians and specialists. It includes a range of subspecialists, specialists, and generalists with overlapping and complementary talents. Some are ``general specialists,'' who fulfill the primary care needs of their specialty patients and often serve as general physicians for other patients. Some healthcare is rendered by nonphysician providers (NPPs) and other healthcare workers. Over the course of time, different factors will influence the need for providers in these various categories, and each will influence the other. Predicting the necessary size of any one component is a difficult task.
In the past, estimates of physician supply and demand did not require particular attention to nonphysician providers such as advanced practice nurses and physician assistants. However, they have become a significant force in medicine, and they will have a strong influence on physician workforce requirements [4]. It is unlikely the healthcare system of the future will demand physician services for what can readily and competently be accomplished by skilled NPPs.
There are approximately 40,000 advanced practice nurses and 25,000 physician assistants practicing today. Although many NPPs work in conjunction with physicians, many do not, and state laws are expanding their authority to practice independently. About half of all advanced practice nurses and 96% of physician assistants are trained in primary care, but only half of the physician assistants remain in primary care. The demand for NPPs is growing, and the number being trained is increasing rapidly. The annual output of these programs is projected to increase from less than 9,000 today to more than 15,000 by the end of the century, with the principal thrust in primary care. Fifteen years from now the number of NPPs will have tripled to almost 180,000, a level equal to one third of the current number of physicians. These trends will have a profound effect on the physician workforce in the next century.
Physician Supply: 1950 to 1992
The physician workforce has tripled in size since World War II. Even on a per capita basis, the number of allopathic and osteopathic physicians has more than doubled over that period of time (Fig 1
). Most of that growth was in specialty medicine, but the primary care workforce grew as well. In 1992, there were 571,000 patient care physicians, 38% of whom were primary care physicians [1, 5], a figure somewhat larger than the 30% figure so often quoted. This is equivalent to 83 primary care physicians and 135 specialists per 100,000 of the population.
|
The second adjustment relates to general specialists. The only really objective data on this were assembled in the Mendenhall study, published in the late 1970s [6]. It found that 60% of the time of specialty internists was spent delivering what we now refer to as primary care, and that number for obstetricians and gynecologists was 80%. The numbers are probably smaller today, but they are appreciable. Although some question whether the skill set of general specialists is adequate to render cost-effective care, the quality of primary care they provide is no different than that provided by primary care physicians. General specialists represent an elastic reservoir of primary care capacity, particularly in the urban setting. To account for their primary care effort, I have attributed 25% of the work effort of ``medical specialists'' and 50% of the effort in obstetrics and gynecology to primary care. The result is a primary care work effort that is approximately 41% of the total. The primary care workforce consists of 82 FTE physicians per 100,000, and there are 125 FTE specialists per 100,000. The total patient care workforce is 207 per 100,000.
Estimating Current and Future Demand
How does this size of the physician workforce compare to the estimate of need? The logic that leads to a definition of current and future need for primary care services differs from the logic underlying the need for specialty services, and I will deal with them separately. However, it should be noted that although the needs for primary care and specialty services can be distinguished, the distinctions between the primary care physicians and specialists who provide these services are significantly blurred, and describing the physician workforce in this polarized way ignores the breadth of knowledge and ability of most practicing physicians. Nonetheless, the national dialogue separates primary care and specialty physicians, and I will follow that pattern.
Primary Care
There has been a tendency to define the need for primary care physicians in relative terms, that is, the fraction of all physicians who are or should be primary care physicians [7]. Fifty years ago, 60% of physicians were primary care physicians. Today, just over one third are primary care physicians. Fifty years from now, the percentage will be different again. However, percent is not the relevant term. Primary care is population based. The principal determinant of need is demographic, and the primary care workforce is best expressed in per capita terms, that is, in terms of physicians per 100,000 of population. When that is done, an interesting fact emerges. The size of the primary care workforce has been relatively constant at about 80 per 100,000 for half a century (see Fig 1
).
How many primary care physicians are needed? We are in the midst of an active, transitional medical market that is moving rapidly to incorporate managed care principles. In fact, it is the managed care sector that has created the greatest numeric demand for primary care physicians. What do managed care systems require? Managed care systems that are efficiently structured and adequately staffed with NPPs require a primary care workforce of 60 to 65 per 100,000 [8]. The fee-for-service sector uses only slightly larger numbers. This suggests that approximately 80 primary care physicians per 100,000 of population should be sufficient today, if they are adequately supplemented with NPPs and efficiently organized.
What should our primary care training goals be for the future? I believe that a somewhat higher target of 85 to 90 per 100,000 is a good first approximation (see Fig 1). This is based on demographic considerations that assume expanded access to care as well as aging of our population, and it considers the significant role that NPPs will play in the future.
Specialists
Although primary care needs can be estimated in per capita terms, the same is not true for specialty medicine. This is because the driving force behind much of specialty medicine is science, and the specialty workforce is largely technology based. Elements of it will expand or contract, depending on how technology advances and how the market embraces these new advances. Although some technologies replace others, making no further demand on the specialty workforce, most do not, and even those that do frequently enlarge the demand for specialty services. Because it is so diverse and because the technology underlying it is so unpredictable, the appropriate size of the specialty workforce is difficult to relate to any standard or proportional characteristic, such as population.
How does one estimate the demand for specialists? A starting point is the use of specialists by managed care organizations [8]. Staff model health maintenance organizations employ or contract with approximately 85 specialists per 100,000 enrollees. However, health maintenance organizations tend to have younger and healthier patients, and an upward adjustment is needed to generalize their experience to all patients. Moreover, these estimates are for maximally occupied specialists, but it is not clear that all specialists can be, or should be, maximally occupied. In addition, specialists are needed for those enrollees who seek care beyond that provided by their health maintenance organizations. Finally, specialists are required to fill roles outside of the spectrum of most health maintenance organizations, such as staffing state mental hospitals or working in fields such as preventive medicine, public health, and occupational medicine. They, too, must be counted. When all of this is done, a reasonable estimate for the number of specialists needed today is 120 per 100,000 of population.
What about future demand? The guideposts are not well defined. Specialists have been increasing in numbers at a rate of approximately 3 per 100,000 of the population annually for the past 25 years. A somewhat lesser rate would have been ideal. If future needs can be met by an increase in the number of specialists at an annual rate of less than 1 per 100,000, the projected demand will slowly grow from 120 per 100,000 today to approximately 155 per 100,000 in the second quarter of the next century. This seems to be a modest estimate of future need considering the enormous efforts being made to find ways to treat diseases and conditions for which little can be done today, as well as the significant amounts of care that will be required by ``half-cured'' patients whose lives will be prolonged and improved but whose need for care will continue.
How do the earlier estimates of physician supply compare with these projections of demand? It was estimated above that the supply of physicians in 1992, adjusted for the work effort of residents and general specialists, was 82 primary care physicians and 125 specialists per 100,000 of population (see Fig 1). These numbers are remarkably similar to the estimates of current demand (80 primary care physicians and 120 specialists). This suggests that the current workforce should be able to do the job. However, the perception is that it is not working in a coordinated and integrated fashion to deliver cost-effective care. Nor is it reaching all who should be served by it. I agree with those perceptions. However, these are not simply workforce problems. They are system problems.
Access
A significant factor driving the primary care agenda is the persistent lack of access to healthcare in some urban and rural areas. Many who lack access reside within designated Health Professions Shortage Areas. More than 35 million people are affected, two thirds of whom live in rural areas. The social and political imperative to remedy their needs is vast, yet the number of physicians required is proportionately small. It is estimated that 5,000 additional physicians would bring the current Health Professions Shortage Areas to minimum standards [7]. Similar needs are now being met by the network of federally designated community health centers which, collectively, utilize only 2,700 physicians. A doubling or tripling of that number could make a profound difference. But these numbers are almost insignificant in comparison with a physician workforce of almost 600,000 and a primary care workforce of over 200,000.
If these comparatively small needs cannot be satisfied from our current abundant supply of physicians, satisfying them by regulating the size and composition of the physician workforce seems unlikely. Yet, it is to meet the needs in rural America and in our inner cities that the greatest legislative energy exists for increasing the total number of primary care physicians. Clearly, other strategies will be necessary, including greater use of nonphysician providers, broadening of the National Health Service Corps, and possibly even mandatory service by all medical graduates. These all are system solutions.
Costs
A major reason for focusing so much attention on determining the right number of specialists is that they have been seen as the major force driving the volume of medical services and therefore the costs of healthcare [9]. The data underlying this conclusion have certain limitations. However, even if these data are accurate, it is unlikely they indicate what will happen in the future. Physicians increasingly are coming together in capitated systems of practice that are more highly organized and use fewer resources. Quality improvement and outcome measures are becoming prevalent, and practice guidelines are being developed. The practice of medicine is changing in this way not simply because of economic imperatives, but because the technologic and social demands on medicine have become too complex to be achieved except within collaborative frameworks. As a result, volume of service is changing from being physician directed to being system directed.
One corollary of this is that the implied guarantee of full employment that physicians have had through their ability to control volume of service will be lost. A second corollary is that if the goal is to decrease healthcare costs, changing the composition of the physician workforce does not seem to be the appropriate strategy. Rather, it seems desirable to train a highly skilled workforce, but to assure the physicians who compose it are able to practice collaborative, cost-effective medicine. Thus, as was the case with access, cost is a system problem, not a workforce problem.
Future Physician Supply
What will happen to the physician workforce in the 21st century? Figure 2 depicts the growth of the total number of patient care physicians over the next 40 years. The solid curve is a model my colleagues and I constructed based on data from 1992 and used in other of our analyses [1, 2]. In that year, the number of first-year residents was 133% of the number of US medical and osteopathic graduates. This curve projects that the total number of patient care physicians will increase to about 850,000 in 2022 and plateau thereafter. A similar curve was constructed by the Bureau of Health Professions in 1990, when the number of postgraduate year 1 residents was 130% of the number of US graduates. However, there has been a steep increase in the number of international medical graduates over the past few years, and both our estimate and that of the Bureau of Health Professions are less than the curve based on the number of international medical graduates in 1994. Postgraduate year 1 positions in 1994 were almost 140% of the number of US graduates. Finally, a curve has been constructed assuming that the number of international medical graduates will be constrained to 110% of the number of US graduates, as proposed by the Council on Graduate Medical Education [3] and included in the Mitchell Health Care Reform bill debated on the Senate floor.
|
Understanding how these projections of physician supply translate into physicians per capita requires an understanding of what is likely to happen to the US population over the next 40 years, a time during which world population will grow to 8.5 billion, twice the population that existed when I was an intern in 1961. Over the past 4 to 5 years, the US Bureau of Census has progressively increased its projections of the future US population. The data regarding the future physician workforce familiar to most students of this subject are those constructed by the Bureau of Health Professions and used by the Council on Graduate Medical Education [3] and others [7]. Unfortunately, the population projections incorporated into their calculations were those issued by the census bureau in 1990 and are substantially lower than those in current use.
Applying these various population projections to our model of the projected number of physicians yields widely divergent estimates of the number of physicians per capita (Fig 3). The 1990 census data used by the Bureau of Health Professions projects a progressive increase in the number of physicians per capita. In contrast, recent estimates of the census bureau create a ``turn-of-the-century bulge'' followed by a return to levels similar to those that exist today. The dynamics that will create this bulge are all operative today, and there is little that can influence it. Everything is in motion already. The training decisions we make today will not have a significant effect until 15 to 20 years from now, just as the decisions of the 1970s are having their major effects now.
|
What will happen to the primary care and specialty workforces in the next century? Figure 4 presents two scenarios based on the proportions of medical students choosing primary care and specialty medicine in the year 2000 and beyond. The first is an increase from the current level of 25% to a new level of 33% of students entering primary care practice, with a reciprocal decrease from 75% to 67% entering specialty practice. The second is a 50:50 mix of primary care and specialty medicine, known as the 50% solution. These projections all use the 1992 resident-adjusted work effort figures presented earlier and the census bureau's 1993 projection of population, and they assume 25% of the work effort of medical specialists and 50% of the effort in obstetrics and gynecology will be devoted to primary care.
|
Two points are important to discern from the first decade depicted in Figure 4. The first concerns primary care. No matter what medical students choose to do, new graduates will not appreciably increase the supply of primary care physicians over the next 10 years. But the market cannot wait that long. What will it do? It is doing it already. Compensation for primary care has increased substantially over the last several years, and some specialists are changing careers or shifting their patient mix to include more primary care. The market is also creating more efficient modes of practice. Probably the most profound consequence of market forces is the increased production of NPPs, many of whom are being given more independent prerogatives. All of these measures are more rapid and more predictable than the training of new primary care physicians, and all will decrease the future demand for new primary care physicians.
The second point concerns specialists. The number of specialists will increase by approximately 20 per 100,000 over the next decade regardless of what choices medical students make today because the pipeline is already in place. It appears inevitable a surplus of specialists will exist during the turn-of-the-century bulge. What will happen? Some specialists may leave clinical practice or reduce their clinical load. Others will practice more primary care, and some without the skills to do so will attain them. This reservoir of underutilized specialists early in the next century represents the major source of physicians to meet the immediate needs in primary care.
The 33:67 Distribution
If 33% of graduates choose primary care, the number of FTE physicians rendering primary care will slowly rise, largely because the total number of physicians is rising, and it will plateau at about 90 per 100,000 early in the next century. This seems ideal for the healthcare environment that is developing. Decreasing the percentage of graduates entering the specialties to 67% will lead to a plateau of about 155 specialists per 100,000, a level that approximates the need for specialists in the second and third decades of the next century. Thus, both the need for primary care physicians and for specialists will be more than met if 33% of medical graduates choose primary care and 67% choose specialty medicine. In view of the growing enthusiasm for primary care among current medical students, it does not appear that regulatory action would be required to achieve this distribution.
The 50% Solution
If, as a result of state and federal mandates, the number of students entering primary care increases to 50% by the year 2000, the number of physicians providing primary care will grow progressively to almost 120 per 100,000 in 2032. Coupled with the simultaneous growth in the number of NPPs, most of whom will be providing primary care services, a serious oversupply will exist, unless primary care physicians begin to specialize as is beginning to happen now to a limited degree.
The number of specialists also will increase initially, peaking at approximately 145 per 100,000 in the first decade of the next century. Thereafter, it will fall progressively to levels not experienced since the 1980s, but at a time when science and technology will have progressed enormously and the need for physicians with special expertise will be substantial. This decrease in the specialty workforce in the next century does not seem to be consistent with the needs and opportunities that will exist in a technologically advanced healthcare system.
Some may see this profound shift toward primary care and away from specialty medicine not as a distortion, but as a desirable change in the characteristics of the physician workforce: a return to a less complicated form of medicine; a move away from research and technology. However, it is unlikely the American public will find such a shift to be desirable. Moreover, it seems unlikely these profound distortions will even occur. In an era of physician abundance, as will be experienced over the next several decades, it is unlikely that a vast excess of primary care physicians would continue if there were needs in specialty medicine, or, conversely, that a shortage of primary care physicians would go unremedied if there were a surplus of specialists. In such an era, surplus primary care physicians will obtain the training necessary for them to provide the first level of specialty care, just as surplus specialists will gravitate to provide continuing and coordinated care for patients with chronic diseases, a characteristic of primary care. Broadly trained physicians, whether generalists or specialists, will be equally comfortable in this domain of middle care. It is incumbent on us to assure that the students we educate have sufficient generalist skills to occupy this domain and to move among the roles they may need to play over the course of their clinical careers.
Conclusion
There are six core issues that emerged from this analysis:
As physicians we are taught to know what we know and to know what we do not know. There is a great deal about the dynamics of the physician workforce we do not know. We also are taught to do no harm. That is a message that must be given to legislators, as well. I believe regulatory intervention into the physician workforce would do harm [2]. There is a better path to assuring a skilled and balanced physician workforce for the future. It entails a continued commitment to excellence and expertise among physicians, together with a commitment to a broad and general education for those entering the profession of medicine. That is the path I believe we need to follow.
Footnotes
Presented at Preparing Your Practice for Change: Thoracic Surgery Into the Next Decade, Atlanta, GA, Sep 24–25, 1994.
Address reprint requests to Dr Cooper, Health Policy Institute, Medical College of Wisconsin, 8701 Watertown Plank Rd, Milwaukee, WI 53226.
References
This article has been cited by other articles:
|
|
 |
|
  R. J. Shemin, S. W. Dziuban, L. R. Kaiser, J. E. Lowe, W. C. Nugent, M. C. Oz, D. A. Turney, and J. K. Wallace Thoracic surgery workforce: snapshot at the end of the twentieth century and implications for the new millennium Ann. Thorac. Surg., June 1, 2002; 73(6): 2014 - 2032. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
