I came to this topic from a physician perspective, wondering what it was that we really needed to accomplish in terms of physician supply. In thinking about it and in addressing this issue at a time earlier in the 1990s when there was a building consensus that we were headed for a massive surplus of physicians, I came to think about the fact that, as a physician, I relied heavily on other colleagues. I am a hematologist/oncologist and I could not practice without oncology clinical nurse specialists. We do not practice alone. We depend on others for what we do. The universe of medical care is a stage that we share with others. The dynamics affecting others are very important in shaping our own practices — as well as in shaping any concept of what future requirements there are for future physicians and what other issues, dynamics, and trends influence this future vision. Where we are going? What will get us there?

It is not necessarily what we all decide in this room. There is a dynamic that moves our society, that moves our economy, and that moves our culture that really is quite apart from day to day decisions — although it includes the individual decisions and thoughts of all of us and the collective thoughts of us in groups. Can one distill these trends down to something that gives a measurable view of the future? That is the point of departure for the comments I would like to make about non-physician clinicians, their relationship to physicians and their impact on the necessary size and composition of the physician workforce in the future.

You were all here yesterday, so you did not have a chance to read this week’s Journal of the American Medical Association. I was not here and I was able read the latest issue, which has a very interesting commentary by the economist Uwe Reinhardt. I’d like to include some of his thoughts as a way to start this discussion. I title it, “Getting the Story Straight.” Reinhardt said, “Although I remain unconvinced that the nation suffers from a huge physician surplus, that thesis remains a conventional wisdom in the policy arena. If it were truly so then it would be difficult to argue before the U.S. Congress that the production of this surplus human capital warrants continued generous public subsidies. Getting the story straight on the physician surplus should have a high place in academic medicine’s research agenda.”

I would like to devote the next 20 minutes or so to trying to get it straight and, in particular, getting straight the role that non-physician clinicians will play. As one looks at the dynamics that fuel this system and that fuels the need for physicians, it turns out that there are four:

  • the economy is a very strong dynamic

  • work effort of physicians

  • demographics of the population generally

  • the substitution for physicians by non-physician clinicians and the substitution among specialists for each other

Let’s just spend a moment on some of the issues aside from the non-physician clinicians, and I will come to that last. First, demographics: It is a boring subject. I am coming to Washington, D.C., again next week to speak to the Federal Forecasters Conference, and there is a whole session dealing with the issues of forecasting the population. We are forecasting how many physicians and non-physician clinicians we need for a future population. But what will be the size of the future population? Here are the predictions of what the population would be in the year 2000 (Figure 1). In 1977, the index year that was used in the Graduate Medical Education National Advisory Committee (GMENAC) Report, the prediction for the year 2000 population was just under 260 million, and so this was used by GMENAC to determine how many physicians would be required to serve this population. Thirteen years later the Census Bureau revised its estimates up to 270 million; and 10 years later up to 275 million; and then, after the year 2000 census was reported, it was increased to 282 million. That still excludes 4 to 5 million who were undercounted. So the real population in 2000 was about 286 million, which is 26 million more than GMENAC considered and 16 million more than Council on Graduate Medical Education (COGME) considered in its landmark reports issued in the early 1990s. Similarly, in 1990, the Census Bureau estimates for the year 2020 were 50 million lower than is likely to be the case, imposing significant errors on the projections of physician need for 2020 made by COGME and most others. So getting the future population right is very important.

Figure 1:

Population Projections

Figure 1:

Population Projections

Then, of course, there is getting it right in terms of age, and for that the year 2010 is the magic number for two reasons. First, it marks a radical downturn in the number of physicians per capita. Secondly, it marks a radical upturn in the number of citizens who will be over the age of 65. Think about 2010, which is only 10 years from now. We cannot do anything about 2010 in terms of educating health care providers because it takes so long to start an educational program and to educate individuals.

The reason that 2010 is magic for physician supply is because it is 40 years after 1970. What happened in 1970? Medical school class size doubled, and so for these last 40 years, we have been producing more physicians than the number of physicians that have retired. But in 2010, we hit the steady state, and low and behold, although we do not produce more physicians, the mothers of this nation produce more babies. Therefore, the population continues to grow while the physician population is static, and in population terms the number of physicians per capita begins to decline. Of course, that coincides with the rising number of individuals over the age of 65. In addition, the physicians themselves are aging because the cohort of young physicians from the 1970s and 1980s is now older, and older physicians tend to work fewer hours than younger physicians.

Another factor to consider is the changing work effort of physicians. We have more women in the workforce. Women physicians work fewer hours than male physicians. There also is a lifestyle thing that is happening. I still do not seem to understand it, because I work a thousand hours a week because I do not know how to do anything else. When you are on every other night and every other weekend as a resident, you think that is what you are supposed to do for the rest of your life. When you are on every fourth night or once a week, the expectations are different. The expectations of a whole age group of which my children are a part are very different than what my expectations were. So we will not be seeing the same work output from future physicians. It may be better for family life, but it has an impact on the number of physicians that we will need. Of course, we know that employed physicians tend to work fewer hours than self-employed physicians, and more are becoming employees, so the whole curve shifts down in terms of effective effort (Figure 2).

Figure 2:

Work Effort Trends and the Supply of Physicians Per Capita 1990–2020

Figure 2:

Work Effort Trends and the Supply of Physicians Per Capita 1990–2020

The economy is another thing that we have to get right. It is very interesting. In an economic analysis that compares the gross domestic product per capita (GDP) and health expenditures per capita, both rise over time in a fairly parallel way (Figure 3). However, as the GDP approximately doubles, health expenditures per capita approximately quadruple. There is a multiplier effect and health expenditures become a larger and larger portion of the total economy. It does not surprise me, and it should not surprise you, because as one has more resources one wishes to expend more of those resources on health. One does not spend more on food; one does not spend more on transportation; one spends more on leisure, health and electronics — on things that one always would want. As a consequence, there is a very strong correlation between the GDP and health employment (Figure 4). The richer the nation, the more people are employed in health care.

Figure 3:

Gross Domestic Product vs Health Care Expenditures 1960–1997

Figure 3:

Gross Domestic Product vs Health Care Expenditures 1960–1997

Figure 4:

Gross Domestic Product vs Health Care Expenditures 1960–1993

Figure 4:

Gross Domestic Product vs Health Care Expenditures 1960–1993

Who are those people? They used to be doctors. Back in the year 1850 most of the health care workforce was comprised of doctors (Figure 5). But throughout the 20th century, other health care workers have entered and taken on many of the tasks that physicians previously did, or taken on new tasks that make it possible for physicians to do what they now do. These include pharmacists, registered nurses, licensed practical nurses, therapists of various kinds, and technicians. In the 1920s a new group appeared, the practitioners who we now call non-physician clinicians (NPCs). This group is enlarging quite rapidly. As a result of all of this, physicians as a percent of the health care workforce have fallen progressively to about 7 percent.

Figure 5:

Health Employment per 100,000 of Population 1850–2000

Figure 5:

Health Employment per 100,000 of Population 1850–2000

Nonetheless, there is a very interesting correlation between the gross domestic product and the number of physicians per capita (Figure 6). This graph goes back to 1929. In the 1950s and 1960s, there was a physician shortage. It was following that period that new medical schools were opened. The fact of that shortage is revealed by the trend line. The data came back to the line in about 1985 and then a little bit left of the line, indicating a small surplus in the early 1990s. They then returned to the trend line in 1999. If one extrapolates that trend out to the future, based on some reasonable projection of the gross domestic product and the known output of physicians, the curves appear as depicted in the top right hand corner of the graphic. As we come into the years after 2010, we will more than replicate the physician shortage of the 1950s. The magic number is 2010. In that year we deviate from a very strong trend line spanning 70 years. When we look at that trend line, it tells us that, on economic grounds, the demand for physicians will far surpass the supply. But, as I indicated at the outset of my remarks, physicians are not the only providers of care.

Figure 6:

GDP per Capita (1996 dollars) vs Active Physician Supply per Capita 1929–1999 and Projected to 2020

Figure 6:

GDP per Capita (1996 dollars) vs Active Physician Supply per Capita 1929–1999 and Projected to 2020

That brings us to the final element of this formula, and that is substitution by NPCs. There are many who contribute to this whole arena. They include nurse practitioners, midwives and physician assistants; they include alternative disciplines, such as chiropractic, naturopathy and acupuncture, psychologists, optometrists, podiatrists, nurse anesthetists, pharmacists and of course dentists. Each has a scope of practice that partially overlaps the scope of practice of physicians. Over the past decade there has been a significant increase in the number of graduates in many of these disciplines, while the number of physicians being trained did not change very much. Nurse practitioner graduates increased 300 percent. The number of physician assistants being trained more than doubled. It is the same for midwives. Chiropractors, acupuncturists and naturopaths increased substantially. But the numbers of optometrists, podiatrists and nurse anesthetists being trained remained fairly flat and clinical nurse specialists decreased a bit as nurse practitioners increased. These disciplines underwent the kind of growth that physicians underwent in the 1970s, and the wave of an increasing supply is just beginning.

There are currently about 75,000 nurse practitioners, but this will rise by the year 2015 to 175,000 (Figure 7). Chiropractors are rising to numbers that are similar. Physician assistants are not far behind. Acupuncturists are taking off, and there are new programs in Asian medicine. There will soon be doctoral degrees in Asian medicine. The numbers are not as big for nurse midwives, but the patient population is smaller. In terms of obstetricians, this group is substantial. Naturopaths remain a small but interesting and growing group. Collectively, this group is increasing in substantial numbers. As a comparative measure there will be more primary care nurse practitioners and PAs than there will be family doctors in the year 2015. In contrast, podiatrists are having trouble filling their classes at the moment. Nurse anesthetists and optometrists are projected to remain fairly flat in population terms.

Figure 7:

Physicians and Nonphysician Clinicians, Changes in Graduates 1992–1998

Figure 7:

Physicians and Nonphysician Clinicians, Changes in Graduates 1992–1998

Licensure is an important issue because not only are the numbers of NPCs increasing, but their licensed scope of activity is increasing. Figure 8 deals with licensure and autonomy. Nurse practitioners are finally licensed everywhere. Illinois was the last to license them. PAs are finally licensed everywhere. Mississippi was the last state to license PAs. Acupuncturists are licensed in only 40 states, and naturopaths are licensed in only 11 states, but the numbers of states is increasing, and the acupuncture bar will extend into all 51 jurisdictions soon.

Figure 8:

Nonphysician Clinician Supply 1990–2015 Traditional and Alternative NPCs

Figure 8:

Nonphysician Clinician Supply 1990–2015 Traditional and Alternative NPCs

Of equal interest is the degree of independence that nonphysician clinicians may exercise. Almost 20 states allow nurse practitioners to practice entirely independent of physician supervision. Certified Registered Nurse Anesthetists (CRNAs) are the same, as are nurse midwives. Chiropractors are independent in all states. Naturopathy and acupuncture are independently practiced in about two-thirds of the states in which it is licensed.

PAs are special. In terms of practicing under a physician’s delegated authority, PAs always do. In some instances, a physician is required on site. However, in the main, the delegated authority can be at quite a distance. It could be a rural community 100 miles away or more with a physician available by phone and visiting the practice site every two weeks to review charts. It is a delegated authority but often a very arms-length one with a very long arm.

The numbers grow every year as more state legislation enhances the independence of NPCs. So too is the case for prescriptive authority. Figure 9 shows the data for controlled substances. Wisconsin, where I am from, is one state that allows nurse practitioners to prescribe independently. PAs prescribe, but of course, under the delegated authority of the physician to whom they relate. But again, look at the bars in Figure 10. The bars indicate the number of states that allow independent prescriptive authority. The portion of the bar indicating independence and the length of the bar, indicating the total number of states, will grow each year as more practice acts are modified. This is a very active process.

Figure 9:

Licensure and Autonomy

Figure 9:

Licensure and Autonomy

Figure 10:

Prescriptive Privileges for Controlled Substances

Figure 10:

Prescriptive Privileges for Controlled Substances

Reimbursement is another important issue. Medicare reimbursement is available for all practitioners except acupuncture and naturopathy. The Balanced Budget Act of 1997 enhanced the reimbursement for nurse practitioners, PAs, and clinical nurse specialists. The recently proposed HCFA regulations would enhance the reimbursement for nurse anesthetists. Medicare is clearly moving in the direction of enlarging the independent reimbursement of non-physician clinicians and so is the private insurance industry, often through state mandate. In addition, an increasing number of states have mandated that private insurers offer insurance coverage for various practitioner groups, such as nurse practitioners and PAs (Figure 11). Chiropractic is the largest one. They have lobbied strongly for it. Figure 11 also shows the number of states that have Medicaid programs that cover NPCs.

Figure 11:

Medicaid and State-Mandated Insurance

Figure 11:

Medicaid and State-Mandated Insurance

What do NPCs do? The nurse practitioners and PAs can do 60 to 90 percent of office-based primary care, routine specialty procedures and follow-up. Many of you are using them in your offices. Midwives, in many states, not only assist with uncomplicated pregnancies but also treat sexually transmitted diseases in male partners and routinely do newborn care. Chiropractors treat most musculoskeletal disorders. Chiropractors tend to locate in small communities, a distribution that is reciprocal with physicians and often in those communities they are the portal to entry, a concept that permeates chiropractic care. Naturopaths are primary care physicians operating at an office-based level dealing with common disorders. Acupuncture is skewed toward pain management, addiction and psychosomatic diseases, but again, acupuncturists often serve as the portal of entry and certainly complement the treatment that physicians might otherwise give.

Optometrists, in addition to refraction, have moved strongly in the area of therapeutic optometry. I think many of you are dealing with this in your states. Most states now offer certification for therapeutic optometry. Many of the vision centers are now called vision and glaucoma centers, as optometrists undertake more and more eye care that previously was in the realm of physicians. A number of states have considered providing access and providing privileges to do laser surgery. Oklahoma is the only state that has done it so far. A number of state boards of optometry have granted that privilege. Wisconsin was the most recent, but in every instance the courts, or political pressure in the case of Wisconsin, have caused them to rescind that privilege. CRNAs provide most anesthetic indications. In addition, in the specialty area, PAs and NPs do urgent care and a variety of other tasks that physicians also do.

The spectrum in which NPCs operate is principally toward the least complicated end of care, whereas physicians tend to operate at the more complex end of the spectrum of care. But there is a degree of overlap and there are shifting boundaries. There is a very clear blurring of boundaries as this triangle moves up. As this triangle of complexity moves, physicians become involved in more and more complex care.

What does all of this mean in terms of the future supply of “physician services?” The demand that would be generated by the economy, coupled with the falling number of physicians per capita and their declining work effort, predicts a shortage. However, if one adds in NPCs at “substitution ratios” of between a one-third and one-half basis for various disciplines, their incremental supply will lessen the projected shortage of physicians (Figure 12). The combined supply of physicians and NPCs follows the demand line until that magic date of 2010. However, as it turns out, the greatest demand growth is in the specialty arena and the greatest supply is in the primary care area. So the combined supply of primary care physicians and NPCs really exceeds the demand, whereas the combined supply of specialty physicians and NPCs falls short of the demand. We are left with a peculiar disparity of a growing supply of NPCs and a growing shortage of specialty clinicians.

Figure 12:

Physician and Nonphysician Clinicians Supply and Demand 1990–2015

Figure 12:

Physician and Nonphysician Clinicians Supply and Demand 1990–2015

Let me conclude by saying that as I see it, population growth will continue to exceed official estimates. Economic expansion seems likely as long as Alan Greenspan is in the driver’s seat. The demand for advanced clinical services will continue to grow. We have a 70-year trend showing that and we have lots of other data from studies both in the U.S. and internationally that I have not had time to show you that further supports this relationship of health care to the economy. Physician supply per capita will soon decline, and physician work effort will continue to decline. NPCs will provide many of the primary care and lower complexity specialty services currently provided by physicians. What I think we are seeing — and we need to understand this — is that a surplus of primary care clinicians and a shortage of specialists is on the horizon. The challenge is that, at a time when there is a perceived physician surplus overall and a particular perception that there is a surplus of specialists combined with a consensus that there is a shortage of primary care physicians, we need to address future needs.

To quote Uwe Reinhardt again, “We need to get it right.” We need to begin to increase medical school capacity looking at the years beyond 2010. We need to begin to plan for an expansion of specialty training programs, again looking beyond the year 2010. We need to moderate the training of primary care clinicians, recognizing the large group that is in the training pipeline. And we need to take advantage of the training of NPCs by promoting systems that effectively integrate physicians and NPCs in clinical practice. That is the challenge. You obviously came here to meet the challenge, and I know that you are committed to doing so.

This article was reprinted from Medical Licensure in the 21st Century: Symposium Proceedings September 6–7, 2000, Washington, D.C. The symposium brought together leaders from medicine, medical education and government to help define the future of medicine, so that licensure will accurately reflect medical education and practice. You may order the hardcover book by phone by calling (817) 868-4076, or from the FSMB Web site at http://www.fsmb.org (from the home page, select the “Publications” link and then the “Order Form” link).