Is There a Surgeon Shortage?

The perceived physician shortage is much in the news of late. A lot of that is the usual media and public over-reaction, but even allowing for that, we need to face a couple of unpleasant truths. First, we don’t have enough surgeons right now. Second, we aren’t training enough to meet the demand over the next 20 years. Most of what follows will provide evidence for these statements. The present and future shortfalls are especially acute in the essential specialty of general surgery, and this article will focus on that area.

Consider the general surgeon, who was an endangered species only ten years ago. There were fewer and fewer students applying, and surgical specialties were taking over many areas which were traditionally the province of general surgery. Thus, it may be a bit odd that we are now worrying about inadequate numbers today. But there are very good reasons to do so.

There is solid evidence that we need more general surgeons than we have now. We’ve been training around 1,000 general surgeons a year for the past 30 years. During that time, the population has grown by nearly half. That means that we need half again as many general surgeons as we had 30 to 40 years ago.¹

The general surgery workforce is admittedly hard to define. Surgeons move out of practice into retirement, research, or administration. Many subspecialists work as general surgeons part of the time. Estimates of the general surgery workforce range widely, depending on who is doing the calculations, and what assumptions they make. An estimate of between 18,000 and 21,000 is probably accurate. Perhaps more significantly, a great number of those general surgeons are now limiting their practices to only a portion of general surgery

The specialty of general surgery has expanded rapidly into many subspecialty areas, especially over the last 20 years. We have breast surgery, oncologic surgery, pediatric surgery, trauma and acute care surgery, surgical critical care, bariatric surgery, minimally invasive surgery, colo-rectal surgery, and transplantation. All have separate training programs; some are ACGME certified, some are not. Some of these subspecialties represent new procedures and areas of practice, not just a different split of the general surgery workload. Who would have thought, 15 years ago, that 5–10% of the general surgery workforce would be engaged in bariatric surgery? To further deplete the number of general surgeons, residency graduates still elect to enter residencies in the specialties of cardiothoracic and vascular surgery. Cardiothoracic surgery has contracted somewhat, but vascular surgery has expanded notably. The workload for both of these specialties will increase as the population ages. Some general surgery graduates become plastic surgeons, although most plastic surgeons are trained separately. Finally, consider the demands of 1,000 or so hospitals with trauma centers for dedicated trauma and acute care surgeons. That alone will add several thousand general surgery positions to the demand for surgeons. Even today, salaries of $300K to $400K for surgeons are common for acute care surgeons. Whatever the effect of the ongoing health care reform, the law of supply and demand is one aspect of the health care system that Congress won’t be able to change.

There is another problem which is more intractable, and which is going to occupy us for the next couple of decades. The politicians seem to be discovering it, which may make it completely insolvable. This problem is known as the demographic trap. We are entering into a generation-long time of increasing demand for health care services, combined with a decreasing number of people who are trained and able to provide services. The demographic trap will work to make the health care system more stressed, less responsive, and less accessible.

The increasing demands for health care are well-known. I’ll spare you most statistics on Medicare, which are both dismal and boring. Consider only this. The number of people over 65, now 36 million, will double to 72 million by 2030.² And those older people will require progressively more health care as they live longer. So what’s the trap? At the same time as demand is rising, the number of health care professionals relative to the population is declining.

One obvious example is the nursing shortage. Insufficient numbers of young people are enrolling in nursing schools. This is in spite of the development of two-year “associate” schools, which are both shorter and less demanding than the old (and vanished) three-year “diploma” schools. Some two-thirds of new RNs come from associate schools. Only a third graduate from the four-year baccalaureate programs, which were supposed to replace the diploma programs. By 2020, an actual drop of RNs per capita is predicted, which will be a shortage of around 20% from estimated workforce requirements.³

Most observers find a shortage in the overall physician workforce. Historically, there was a great expansion of allopathic (MD) schools between 1960 and 1980, doubling the places available. But, perhaps in reaction to this, there came to be a perception that too much had been done. In the 1980s, several studies predicted a surplus of physicians by 2000. Those making these predictions included GMENAC (Graduate Medical Education National Advisory Committee), COGME (Council on Graduate Medical Education), the Pew Commission, and the Institute of Medicine. All of these were authoritative, all had at least informal governmental backing, and all were conducted by recognized experts. And all of them were wrong.

These studies used a time-based method to project physician needs. To greatly oversimplify their complex and varied methodologies, each of these studies estimated how much physician time would be required for a large number of services. They added up all the services which would be required in a population, assigned a time value to each, and calculated how many physicians would be needed to provide the services. From medical school graduation rates, immigration rates, and estimates on the average years of service for physicians, they could predict the number of physicians available at any period in the future. This approach is well in line with studies of industrial efficiency, but it failed miserably when applied to physician work. The studies all estimated that there would be a surplus in specialties and subspecialties, by the year 2000 or so. That’s exactly the opposite of what happened. But few things are as persistent as a bad idea, especially if it leads to a desired conclusion. Even as it was proving to be wrong, this viewpoint still persisted through the 1990s.⁵ In some poorly informed circles, like Congress, it persists today.

The basic mistake of these predictions was to assume that the U.S. health system would become highly structured along the lines of existing Health Maintenance Organizations (HMOs), such as Kaiser-Permanente, and that physicians would work far more efficiently than they had in past decades. This simply failed to happen. Outside of a few systems, managed care and HMOs obviously failed to produce increased efficiency. Other assumptions were also greatly optimistic, such as the amount of work that each physician would do over a lifetime.

Not everyone was wrong in the 1980s and 1990s. Dr. R.A. Cooper, for one, disagreed with the consensus. He used an entirely different approach, based on actual trends in physician workforce and population growth. He correctly predicted that there would be no physician surplus in 2000, and now predicts a shortage of 200,000 physicians by 2020.⁶ It should be noted that the American Association of Medical Colleges, after recommending no increase in class size in medical schools for 25 years, is now asking for a 30% increase in allopathic medical school graduation rates over the next 10 years.⁷ Osteopathic medical schools, while a much smaller percentage, have been in the process of increasing their enrollment and graduation rates. So have the Caribbean schools, which draw heavily upon students from the United States.

Let me put in a word about the workforce in research. Research PhDs aren’t involved in patient care. However, the outlook is not especially rosy there, either. There is great concern that the scientific workforce doesn’t match the projected demands over the next generation. The age of grant awardees continues to increase, and there is great concern that too few young biomedical research scientists are in the pipeline.

The environment of the health care system over the next two decades will be dominated by shortages of health professionals. We are seeing problems now, and they will become worse. Hospitals, especially in rural or underserved urban areas, will have increasing difficulties getting and retaining staff. The burden of growing shortages will fall most heavily on the poor. There will be significantly more employment of less-trained health workers to either supplement or replace more-skilled professionals. There will be a great need for efficiency, and for better ways of organizing the work. The demographic trap has been sprung, and we are all caught in it.

The second truth, that we aren’t training enough general surgeons, follows directly from the present shortage. We haven’t increased the number of general surgery trainees for 30 years. We’re still graduating about 1,000 general surgeons each year, which was probably enough 30 years ago. It isn’t now. As already noted, a significant number of those move off into other specialties, but some 500 to 700 remain to be added to the workforce each year. The ratio of general surgeons to population was 6.93 in 1974 and 7.1 in 1994, so there seems to have been a reasonable balance up until about 15 years ago. The current figure is around 6.4, depending on how many general surgeons are counted, so it’s evident that the ratio is dropping. But this figure is not entirely comparable to earlier figures, because of increasing sub-specialization. As general surgery fragments into sub-specialties, and the number and diversity of operations increases, we will need relatively more general surgeons than we had 15 or 20 years ago.

The aging population will strongly impact the demand for physicians.⁸ An analysis by Etzonin et al⁹ in 2003 looked at operative rates by age cohort, which allowed them to estimate future operative case loads as a function of the changing demographics of the population. They projected an increase of from 14% to 47% for surgical specialties by 2020, with most fields showing a 30% to 40% increase. The increase clearly depended on how many older patients were having operations. Specialties like thoracic surgery (42%) and ophthalmology (47%) head the list. General surgery was 31%. To be sure, these are only estimates. To assume that operative case loads will increase, one must also assume an increasing number of surgeons. Etzonin’s analysis is more an indication of increasing demand than it is a prediction of the future numbers of operations.

The general surgeons who trained 30 years ago are now at retirement age. We are now at steady state. We can only increase the net number of general surgeons by adding to the number of surgeons trained each year. Currently, we are increasing that number very slowly. We will probably be up to 1,100 per year by the middle of the decade. At that rate, we will continue to fall further behind. We probably should be training 1,500 per year to make up for the shortfall and to allow for future growth in the population and in the workload. We’re not even close.

Why not? There are several reasons. It is not far past that general surgery had difficulty recruiting even 1,000 medical students per year. That is, fortunately, behind us. There are currently around 1,500 applicants for those 1,000 places, including both American and international graduates, although not all are suitable candidates. Then, too, it’s difficult and expensive to train surgeons, and there is no great rush of hospitals who wish to establish new training programs. While there are new medical schools starting up, most of them are built around existing teaching hospitals, which already have training programs. Finally, there is a cap on the total number of trainees which Medicare will support. Any chief of surgery who wishes to expand a training program must persuade someone else to drop a residency position, or talk the hospital administration to paying for the additional residents “above the cap.” Congress may lift the cap, of course. In a few years. Reluctantly.

So. The blunt fact is, there will be a real problem going into the future. Rural hospitals already have a terribly difficult time attracting general surgeons, and that will become worse as many older surgeons retire. Urban hospitals will see their general surgeons splitting up into sub-specialties, with fewer and fewer “old-time” general surgeons available to take call, handle trauma, and care for the poor. It is unlikely that reimbursement will increase from either insurance or government, which means that the private practice of general surgery will be increasingly uneconomic. So, hospitals will be forced to hire enough general surgeons to carry the workload. Even today, general surgery is becoming more and more a hospital-employed specialty. To the anguish of many, and the inconvenience of all, that tendency will continue. There may eventually be an increase in the numbers of surgeons being trained, but that won’t happen quickly.

How much can this shortage of general surgeons be generalized to the rest of the surgical workforce? Sadly, many other specialties are predicting dismal future. Cardiothoracic surgery, for example, is now training fewer surgeons than it was 40 years ago, at the beginning of the great increase in open heart surgery. Other specialties are in better condition, but no one is optimistic. Perhaps there will be some benefits to the shortages. After all, having an under-employed surgeon standing around is about as non-productive as one can imagine. Forcing specialty operations to be concentrated in fewer hospitals may actually improve care, as long as it is not carried to the point of creating barriers to access. Nonetheless, it’s clear that the situation is not going to improve, even over the next decade or so.

So, to all the other physicians out there, let me conclude with this thought. Be very kind to your local surgeon. If he or she leaves, you may not find another.

Biography

Charles W. Van Way, III, MD, MSMA member since 1989, is a Professor of Surgery and the Sosland/Missouri Endowed Chair of Trauma Services at the University of Missouri-Kansas City School of Medicine.

Contact: vanwayc@umkc.edu