Private sector recommendations and federal legislation affecting recent growth3

Much of the expansion in physician supply in the late 1960s and early 1970s can be traced to concern expressed as early as 1959 that the future supply of physicians would not be adequate for an increasing population unless new medical schools were constructed and the numbers of students in existing schools were increased.⁴ However, several years elapsed before the recommendations were buttressed by federal legislation—the Health Professions Educational Assistance Act of 1963, which provided funds for construction.

Further alarms of an impending shortage of physicians were sounded in 1964 in the Coggeshall report to the Association of American Medical Colleges (AAMC). The report concluded that “more physicians must be trained as quickly as possible…. It must be recognized, however, that it is not likely that America will ever be able to produce all the physicians that the nation would like to have.”⁵

The concept of shortage was reiterated in 1967 in the report of the President's National Advisory Committee on Health Manpower. It was recommended that “production of physicians…be increased beyond presently planned levels by a substantial expansion in the capacity of existing medical schools and by continued development of new schools.”⁶

Both the American Medical Association (AMA) and the AAMC endorsed this policy in a joint statement released in February 1968 and agreed “that all medical schools should now accept as a goal the expansion of their collective enrollments to a level that permits all qualified applicants to be admitted.”⁷ Soon thereafter a health manpower act was passed by Congress; it provided loan and scholarship money, as well as funds for construction and operating costs of medical schools. Eligibility for funds was linked to a requirement to increase the school's class size (capitation grants).

Medical school enrollments

As a result of the private sector consensus and the infusion of government funds, the number of U.S. medical schools increased in number from 89 in 1965 to 127 in 1982 (Table 1). Enrollments of first-year students nearly doubled over this period from about 9,000 in 1966 to more than 17,000 in 1982 (Table 1 and Chapter 7, Table 2). Osteopathic school enrollments have grown at similar rates (Table 2).

TABLE 1

Enrollment of First-year Medical Students, 1965 to 1982, Allopathic Schools.

TABLE 2

Enrollment of First-year Medical Students in Osteopathic Schools, 1965–1982.

Enrollments in allopathic schools have fallen off slightly, as of the 1982/83 academic year, and are expected to remain at this reduced level or to continue a modest decline in the next few years. However, because of the time lag in training, the number of graduating medical students will continue to rise through 1985, and the number of new entrants to medical practice will not peak until about 1990.

Inflow of Foreign Medical School Graduates

The annual inflows of FMGs have been very uneven, reflecting changes in health manpower policies, and immigration and licensing policies (Table 3). As a group, FMGs represented nearly one-third of the additions to total physician supply between 1966 and 1976.¹ In 1981/82, FMGs accounted for 25 percent of all residents in first-year postgraduate training and 19 percent of all residents (Table 4). Nearly half of all FMG residents were United States citizens who studied abroad. The recent increase in numbers of FMGs is due to the increase in U.S. FMGs (Table 4). It is now believed that a largely new category of FMG, the unemployed European doctor, is attempting to enter the physician work force in the United States.⁸

TABLE 3

Numbers of Initial Licenses Issued by State Boards of Medical Examiners, Selected Years 1960 to 1981.

TABLE 4

Foreign Medical Graduate Participation in Graduate Medical Education, 1979 to 1981.

Projections of future supply

Projected increases in physician supply are detailed in two recent studies: the 1980 report of the Graduate Medical Education National Advisory Committee (GMENAC),⁹ and the 1982 report to Congress by the Bureau of Health Professions (BHPr) of the Department of Health and Human Services.¹

The GMENAC report estimated that the supply of physicians would reach 536,000 by the year 1990, and 643,000 by the year 2000 (Table 5). The total supply figures were projected using a mathematical model that started with the present supply and (1) adjusted for attrition through death, disability, and retirement; and (2) adjusted for additions of United States medical school graduates, foreign medical graduates, and residents in training.

TABLE 5

Physician Supply Projected by GMENAC, and U.S. Population, 1978, 1990, and 2000.

The GMENAC projections assumed that numbers of U.S. medical students entering allopathic schools would increase 2.5 percent per year between 1978/79 and 1981/82 and then remain constant at 18,151; that numbers entering osteopathic schools would increase 4.6 percent per year through 1987 and then remain constant at 1,868; and that numbers of FMGs entering residencies in the United States would increase to 4,100 by 1983/84 and then remain constant. Residents in training were included in the supply figures as 0.35 full-time practice equivalent.

We now know that enrollments through the remainder of the 1980s will probably be lower than GMENAC expected (e.g., in 1982/83, first-year enrollments were 17,254 rather than the 18,151 projected). However, the current numbers of entering FMGs are somewhat higher than expected—4,500 versus 4,100. Thus, in balance, the GMENAC estimates of total future physician supply are not too different from what they would be if the calculations were repeated today.

The Bureau of Health Professions more recent report estimates that the total supply of M.D. and D.O. (Doctor of Ostepathy) practitioners will climb to 591,200 by the year 1990, representing a 32 percent increase over 1980 (Table 6). The BHPr projections are higher than those of GMENAC, since they count residents as full-time equivalents of practicing physicians, while GMENAC included residents at the rate of 0.35 their number.

TABLE 6

Physician Supply Projected by Bureau of Health Professions, and U.S. Population, 1980, 1990, and 2000.

Whatever the projection methodology, it is clear that the pool of physicians in the United States has grown sharply and will continue to grow into the 1990s. We know reasonably well how many doctors the nation will have in 1990. What their geographic and specialty distribution will be and how many will be required to meet the health needs of the population, and consequently how many should be trained, is much less certain, as discussed below.

Supply of Nonphysician Providers

The increase in numbers of practicing physicians has been accompanied by a parallel rise in the supply of nonphysician health care providers. Between 1965 and 1980, the pool of registered nurses (RNs) increased by more than 100 percent (Table 7). This same period witnessed the emergence of two new categories of health professionals-nurse practitioners (NPs) and physician assistants (PAs).

TABLE 7

Growth in Numbers of Active Health Professionals, 1965 to 1980.

The first PA program was started in 1966 at Duke University. The role envisioned for the PA was aiding in many of the primary care responsibilities that burdened the already overworked physician and substituting for physicians in shortage areas.¹⁰ The first NP training program was introduced at the University of Colorado at about the same time to give pediatric nurses expanded clinical training for primary health care of children. Subsequent programs across the country prepared family nurse practitioners to meet the primary health needs of all age groups, geriatric nurse practitioners to care for the elderly, and certified nurse midwives to care for low-risk pregnant women.¹¹

As of 1980, there were more than 1,272,000 active RNs in the U.S., approximately 17,000 NPs and nurse midwives, and an estimated 8,800 active PAs (Table 7).

In many cases the services of nonphysician providers overlap those of physicians. Their work is not only auxiliary to that provided by the physician, but sometimes it substitutes for care otherwise provided by physicians. Increasingly, nonphysician providers and physicians are competing against each other for patient care services. For this reason, any comprehensive consideration of physician manpower must be integrated with manpower assessments of all other health professionals.

Nonphysician health care providers

According to data from the AMA Socioeconomic Monitoring System, NPs and PAs working in physician's offices increase physician productivity about 20 percent. In 1981, there were 3.9 vs. 3.2 patient visits per week respectively for physicians who did vs. did not employ NPs and PAs.¹⁸

Further uncertainty is related to the share of health care services that would be assumed by the growing number of health care providers in competition with physicians, e.g., nurse midwives competing with obstetricians, NPs with primary care physicians, psychologists with psychiatrists, optometrists with ophthalmologists, and so forth. Traditionally, physicians have been able to maintain their professional and economic position in competition with other groups.

Women Physicians

In 1981, 12 percent of American physicians were women,¹⁹ and women were 31 percent of the entering classes in U.S. medical schools.²⁰ It is expected that they will be 16 and 20 percent of the total pool of physicians by 1990 and 2000, respectively.¹

The increased presence of women in the profession complicates estimates of requirements because women physicians, at least in the past, have worked somewhat fewer practice hours per week^*. Women also now have a wider choice of specialties than they did in the past, which will modify their overall practice characteristics.

Physicians in Research and Academia

In estimating physician manpower requirements it must be borne in mind that about 10 percent of physicians pursue careers in academia,^** teaching and carrying out research either in conjunction with clinical practice or instead of it. According to a report from the National Academy of Sciences,

labor market data show that personnel shortages continue in the clinical science fields, primarily because of the strong demand in medical schools for faculty members to conduct research as well as teach and provide patient care, and the difficulty in recruiting young physicians for research careers…. There are signs of improvement in market balance, but conflicting trends make interpretation difficult.²²

In 1981/82 there were 2,264 budgeted but vacant full-time faculty positions in clinical departments of U.S. medical schools, and 668 such positions in basic sciences departments.

Several indicators suggest a trend of decreased participation by MDs in research—what Dr. Donald Frederickson, former Director of NIH, has called the “dwindling bedside connection.”²³ There has been an overall decrease in numbers of NIH trainees and fellows; in addition, the percentage of NIH postdoctoral trainees who are MDs dropped from 59 percent in 1965 to 30 percent in 1980.²⁴ The number of competing grants awarded to MDs by NIH has increased (1,010 in 1973; 1,489 in 1978), but the percentage of awards that are to MDs has decreased (35 percent in 1973; 27 percent in 1978).²²

On the other hand, students continue to express an intention to pursue careers in research.^*** Although there has been a drop from the 1960 value of 39 percent, more recently (1978, 1979, 1980) a steady 21 percent of graduating MDs indicated they intend to have careers as researchers.²⁵

Several factors lead to uncertainty in estimating future requirements. The rate of increase for the budget of the National Institutes of Health has dropped markedly—so that shifts away from training programs and contract research have been necessary to maintain the independent investigator initiated (R01) grant program at desired levels;²⁶ shifting budget allocations create uncertainty as to how much money will be available in the future to support faculty research. The total enrollments in medical schools probably will decline very gradually in future years, suggesting possible decreased need for faculty. But there also is a growing demand for clinicians to participate in faculty practice plans, which are becoming ever more important in the economy of medical schools (Table 1, Chapter 8). Most analyses project a need for clinical investigators and medical school faculty that will continue to outstrip supply.^22,24,25

Minority Physicians

Another important consideration in estimating physician requirements—and deriving manpower policy from that estimate—is the role in the health care system of physicians who are members of minority groups. A period of increasing enrollment (1968 to 1974) was followed by a ten-year period of more or less uniform first-year enrollment by underrepresented minorities at a level of about 9 percent (Table 2, Chapter 7). The changing financing of medical education raises questions about future supply of minority physicians (Chapter 8 and Appendixes C and D).

A number of reasons have been put forward for the particular need to assure adequate participation in medical education by minority persons. These include differences in language or diet, for example, which make provision of optimal health care across cultures difficult; the greater percentage of black MD graduates who express an intention to practice in traditionally medically underserved areas²⁷ also is important to future policy planning (see Appendix D).

The Physicians

The majority of physicians in the United States work in private practice and on a fee-for-service basis.²¹ Increasingly, however, physicians are expressing a preference for institutional practice in which the advantages include regular hours, protection from the most demanding elements of practice, fast start-up at no personal expense, group malpractice coverage, assured salary, and protection from competition. Institutional practice opportunities are available in large group practices, Health Maintenance Organizations (HMOs), hospitals, incorporated systems for providing emergency services, the Veterans Administration, the military system, and elsewhere. (See Chapter 4.) In these arrangements physicians find advantage in organizing as employee groups for purposes of collective bargaining with employers on issues of salary, hours of work, and fringe benefits. A labor-management mentality develops, which may perhaps conflict with the professional ethos.

The Facilities

Substantial changes are occurring in the corporate structure of the nation's 7,000 hospitals. About one-third of the hospitals have formed multihospital systems (horizontal integration) wherein they experience advantages in marketing services, savings through bulk purchasing, and improved access to markets, thus improving their ability to purchase equipment, to add beds, facilities, and services, and to renovate. The largest of these multihospital chains are for profit (Hospital Corporation of America, Humana, American Medical International), with for-profit beds now comprising about 11 percent of the national total.²⁹ Corporate restructuring also occurs as hospitals attain ownership of health promotion centers, ambulatory care centers, diagnostic centers, satellite facilities, rehabilitation facilities, and nursing homes (vertical integration).

To the extent that these institutions promote and are able to achieve greater efficiency in health care delivery, requirements for health professionals will theoretically be reduced.

The Patients and Payer-Underwriters

Until about the middle of this century, the patient generally paid for health services out of pocket via direct transactions with the patient and the doctor, the pharmacist, or the hospital cashier. The advent of health insurance and third-party payers, however, has distanced the patients from the financial consequences of their transactions in the medical marketplace. In between the money providers and the patients has grown an underwriting complex of insurance companies, including Blue Cross, Blue Shield, Prudential, Aetna, and many others, whose function is to distribute the financial risk over all enrollees and to facilitate the transaction between the patient and the health service provider.

Increasingly in the past, and planned on a larger scale for the future, employees have been offered a choice of combinations of underwriters and providers. These combinations, or plans, compete for employee groups and negotiate with physician groups and hospitals for the broadest services at the lowest price. Some physician groups have entered the competition by establishing a variety of organizational structures and numerous cost-saving arrangements. These include HMOs, IPAs (independent practice associations) and PPOs (preferred provider organizations—which offer services at a cut rate to selected employee groups). The underwriter-provider plans compete with each other by offering different arrays of coverage, deductibles, co-insurance, limits, premiums, physician groups, and hospitals.

If physicians are willing to staff these plans on a large scale, one might envision the incorporation of the health services system. Employers, facing increased pressures to reduce costs, will negotiate with employees for the lowest acceptable health payment package. Employers and government agencies on behalf of employees, retirees, the poor, and others will in turn negotiate with underwriting companies, physicians groups, and hospitals for specified services on the basis of a prearranged annual fee. In each case, rules would govern the operation, many of them designed to reduce utilization and costs. Physicians and physician groups that commit themselves to restraint in the utilization of health services would be in demand by underwriters and employers. Algorithms, agreed upon in advance, would determine the specifics of medical care. All of these cost-saving measures, if effective, would tend to reduce requirements for personnel.

Although an incorporated health services system is intended to manage the business aspects of health care, the result could have a profound effect on the individual interaction between one doctor and one patient, and on the characteristics of the patient's personal health services. Individual patient and individual physician discretion would be reduced under these circumstances. But the realization of this incorporated structure cannot occur without the willingness of physicians to staff these plans in large numbers. The new physicians who are hired will have to abandon professional freedom in the economic sense and accept the restrictions of incorporated employment.

Some Results of an Increasing Supply of Physicians

As discussed above, the number of actively practicing physicians will be increasing sharply in the 1980s. The increasing supply of physicians is already being felt by young residents in training as they search for a place to practice. They are finding many areas already saturated with physicians, the competition for patients very high, and many hospitals unwilling to accept additional doctors on the medical staff.

One result is the increasing favor with which new physicians look upon salaried positions, or assured practices, within an incorporated structure as provided in HMOs; large multispecialty group practices; in hospital ambulatory, emergency, critical care, and subspecialty procedure facilities; with manufacturing concerns; in the military system and Veterans Administration; and in other public institutions.³⁰ The influx of newly trained physicians into practice has been so brisk in the past years, and will be sustained at such high rates through the 1980s, that of all the physicians in practice in 1990, about 40 percent will have entered practice since 1978.

The effects of the rising number of physicians and of their changing attitudes toward employment are evident. Almost 50 percent of all physicians now derive some of their income from salary. The Physician Practice Study documented that institutional practice was the main practice for 15 percent of newly trained family physicians, 42 percent of primary-care-track general internists, 23 percent of general internists, and 31 percent of subspecialty internists.³⁰

A likely scenario follows. The rapid rise in the number of physicians will provide the final impetus to complete the incorporation of health services. The negotiations for health services will be conducted in a buying-selling-servicing mode as in most markets. Like service contracts for automobiles, office machines, and home appliances, the allowable services will be agreed upon in detail in advance. An elaborate set of rules governing physician decisions will be adopted by each physician corporation. These rules will specify the allowable use of certain diagnostic tests for a given set of symptoms; number of visits per year for a given condition; components of the annual physical exam; alternatives to elective surgery; criteria for hospitalization; duration of hospitalization; allowable drugs; limits to therapy of the seriously or terminally ill; the use of preventative, rehabilitative, mental health, and nursing home services; and most other decisions previously delegated by society to an individual physician's judgment in an individual doctor-patient relationship. The major underlying emphasis of the rule making will be on cost reduction.

There will be decisions on which health professionals will provide primary care. In adult medical care, for example, roles will be decided upon for family physicians, general internists, subspecialty internists, obstetrician-gynecologists, and nurse practitioners, for psychiatrists, psychologists, and psychiatric social workers. These sets of rules will in the future make possible a more precise specification of the relative number of generalists, specialists and subspecialists needed.

Many of the changes in the health care system described above have implications for manpower requirements. The growth of HMOs and other prepaid plans, the trend of physicians to select more salaried positions, and the rule making aimed at cost reduction, for example, are expected to lower requirements. These changes will tend to diminish the open-endedness of the market for medical services. Institutions will establish the number of physicians that can be efficiently employed. While the total effect of a transformed health care system on supply and requirements cannot be predicted with certainty, it would be advantageous to both the institutions that train physicians and the residents in training to have available timely information on changing requirements and employment patterns.

Proposed Study

The structure we are recommending will be referred to as the National Health Manpower Study and Policy Development Committee (or the Study). The intent is to develop a continuing, comprehensive, long-term health-manpower program, although in this document only a three-year proposal is being presented. A major report on this first phase of the study would be issued three years after inception of the Study, which could use 1982 baseline data. The Study would employ state-of-the-art manpower methodologies; newer modifications should be introduced to advance the field. Supply and requirements figures would be calculated for five-year intervals 1987, 1992, and 1997. Every effort would be made to assess the data for regional, state, and local areas, not only at the national (aggregate) level.

Development of scenarios that integrate manpower supply with demand and need would be particularly instructive. One function of the Study would be to construct the scenarios, and to identify data needs and methods to collect that data.

It is recommended that the Study have two structural subcomponents: the staff, whose function would be to collect and analyze data; and the committee, whose function would be to interpret the information, weigh alternatives, formulate policy recommendations, and move toward national consensus.

The Committee would be composed of 20 to 30 individuals, drawn from a broad constituency—medical schools, nursing schools, schools of public health, and other health professions schools; professional organizations; underwriting organizations; health economists; political scientists; public administrators; state and federal governments; consumers; and others. The chair would be expected to spend about one-fourth-time on the endeavor.

The staff would be capable of primary data collection, either in-house or by subcontract, and sophisticated analyses. The staff also would take advantage of data collection activities of government and private agencies (BHPr, AAMC, AANC, AMA, etc.). The staff director would be a senior, experienced, and professionally respected individual. A staff of approximately 10 full-time equivalents is envisioned, including an administrative director, senior analysts, statisticians, economists, data-management personnel, and secretaries.

The committee might meet quarterly for two days at a time. Subcommittees or task forces could be organized for specific assignments, each with a staff person. A preliminary report should be prepared by the end of the first 18 months and be widely circulated for comment and criticism. A series of regional conferences should be held at various locations, actively involving in the process state and federal representatives as well as all other interested parties. The objective would be to reach broadly toward all concerned and to develop a broad consensus. A major report should be issued at the end of the three-year period.

Home for the Study

There are several possible homes for such a project. Within the federal government, the Bureau of Health Professions of the Department of Health and Human Services has a mandate to report to the President and Congress on the status of health professions personnel in the United States. The BHPr has a great deal of experience collecting and analyzing such data. In developing manpower policy, however, the federal government would be necessarily tied to political concerns.

Both the American Medical Association and the Association of American Medical Colleges have been actively involved in manpower data collection, assessments, and policies. Neither group is, however, explicitly concerned with health professionals other than M.D.s. The AMA and the AAMC are also viewed as looking predominantly at the needs of the profession, although the needs of the public are also high on their agendas. In determining requirements for health professionals in particular, it is important that the work be carried out on a neutral level, drawing together the views of all sectors.

The Institute of Medicine, which can bring together many disparate interests, is in a good position to provide leadership. Its elected membership spans a broad range of the health professions and includes health policy analysts from the fields of law, economics, and public administration. The Institute of Medicine Committee on National Needs for Biomedical and Behavioral Research Personnel could provide extensive data and analyses for research manpower policy formulation. We believe that the Institute can provide the needed connection between manpower planning and health professions education.