The Impending Shortage and the Estimated Cost of Training the Future Surgical Workforce

The Impending Shortage and the Estimated Cost of Training

Thomas E. Williams, Jr., MD, PhD, FACS; Bhagwan Satiani, MD, MBA, FACS; Andrew Thomas, MD, MBA; E. Christopher Ellison, MD, FACS


Annals of Surgery. 2009;250(4):590-597. 

In This Article


History of the Physician Workforce Debate

The cyclical debate about the future physician workforce has now turned in the direction toward predicting shortages of both primary care and specialist physicians. Unfortunately, several studies since the early 1900s related to physician workforce have often been wide of the mark. The Flexner report in 1910, which reported that too many physicians were being trained, led to a 80% reduction in the number of medical schools and a subsequent drop in per capita physician ratio from 175 to 125 per 100,000 population.[6] Again in 1932, the Commission on Medical Education of the Association of American Medical Colleges suggested a surplus of physicians in the United States.[7]

The Study on Surgical Services for the United States (SOSSUS) was one of the earliest studies jointly conducted by the American Surgical Association and the American College of Surgeons from 1970 through 1975.[8] Using manpower and workload estimates, the committee reported that 1600 to 2000 residents completing surgical training each year between 1976 and 2012 would fulfill the need for surgeons in the United States. The SOSSUS study also resulted in closing down many residency programs not affiliated with academic medical centers. The Graduate Medical Education National Advisory Committee (GMENAC) reported in 1980 that the country was likely to experience a surplus of 144,700 physicians by the year 2000.[9] The GMENAC assumptions were based upon experts estimating the amount of work required by physicians to provide specific services for medical and surgical problems. The 1994 annual survey by the American College of Surgeons of residents in surgery seemed to indicate that over 4000 surgeons completing core surgical residency (those offering primary certification by an American Board of Medical Specialties [ABMS] board) training per year would be sufficient to take care of the surgical needs of the country.[10]

The 1990s was a period of dramatic change with the seeming dominance of managed care as the prevailing model of healthcare for years to come. Many professional and health policy think tanks issued their estimates on manpower needs based on the contraction of services forced by managed care. The Council on Graduate Medical Education, a quasi-governmental group authorized by Congress, along with many other groups including the Institute of Medicine, the American Medical Association, the American Osteopathic Association, and Association of American Medical Colleges, reported to Congress their consensus that a surplus of physicians was on the horizon.[11] Indeed, the PEW Commission estimated that 80–90% of the population would be receiving their healthcare through managed care or similar systems leading to a surplus of physicians estimated at 100,000 to 150,000 and, therefore, went so far as to recommend reducing the entering medical school size by 20% to 25%.[12] With the projected surplus of physicians and budgetary pressures, funding for GME was frozen at 1996 levels with the passage of the BBA of 1997.

Starting as early the mid 1990s, warnings of shortages in several specialties including radiology, anesthesiology, and cardiology were published.[13,14,15] One of the skeptics was Richard Cooper who used an economic trend model based on Gross Domestic Product (GDP) per capita to estimate demand for physician's services.[16] He followed this work with a series of other articles that convinced others to reexamine previous estimates of the workforce. Eventually that effort led to the Association of American Medical Colleges recent recommendation to increase the enrollment of first year medical classes by thirty per cent.[17]

Methods of Workforce Estimation

There is often a philosophical disagreement between those who believe that if the healthcare system decreases the number of physicians it will suppress demand and therefore decrease costs and the opposing view that artificially curtailing demand is harmful and will result in long-term adverse consequences. Regardless, for policymakers on both sides it is important that they have reasonable forecasts about the future supply of surgeons, needs of the population and the important variables that influence these forecasts. With regard to estimating the future surgical workforce, there are several methods to calculate physician supply and demand. The work effort analysis reported by Etzioni et al involves calculating surgical workloads with relative value units, Diagnosis Related Group's or numbers of procedure.[18] The second technique is the inventory method or the Physician Supply Model and the Physician Requirement Model (PRM) used by the Health Resources and Services Administration of the U.S Department of Health and Human Services.[19] This supply model tracks physician supply in 36 medical specialties by age, gender, country of medical education, and patient care activity and estimates the number of active, full-time equivalent physicians. The third method is the economic determinant method of Dr. Cooper in which the demand for medical and surgical services is correlated with GDP per capita and population.[4,20]

We have used a simplified population model that starts with the number of practicing physicians in the preceding year, adds new entrants (both US medical graduates and IMG's), subtracts the attrition due to death, disability and retirement and then arrives at the number of active physicians in the workforce. Like the Physician Supply Model and Physician Requirement Model, we also assumed that that the existing pattern of population growth, insurance coverage and the economic state will not change. In addition, we have assumed that the training paradigm will not shift, that Surgeons will continue to practice for thirty years after residency and that their practice patterns will not change. Despite the methodological differences between our study and that of Cooper, our estimates of shortages are within 5% of each other. In our analysis we predicted a shortage of 203,560 physicians by 2020 and Cooper predicted a 200,000 shortage by the year 2025.[21]

Additional Variables Impacting Workforce Estimations

We believe our estimates are fairly conservative since most of the variables mentioned are likely to decrease the supply of surgeons and increase demand for surgical care. For instance, aging of the physician workforce and early retirement are important variables, which could lead to more retirements than estimated. Physicians over the age of 55 constitute 35.78% of US physicians in 2007 compared with 27.52% in 1985.[22] Peak working hours are attained by physicians in the 50 to 54 age group, after which physicians tend to work less.[23] In a recent survey of over 270,000 physicians (50,000 specialists) by The Physicians Foundation and Merritt, Hawkins and Associates, 49% said they plan to reduce the number of patients they see or cease practicing.[24] Twenty percent of physicians stated they would cut back, 13% would seek a job in a nonclinical setting, 11% planned to retire, and 10% intended to work part-time. An earlier study by Jonasson and Kwakawa of GS fellows of the American College of surgeons showed a rise in the average age of retirement from 60.45 in 1984 to 62.97 in 1995,[25] but a more recent report indicates that in 2000 the retirement age was down to 58.[26]

Another important variable is the number of women physicians choosing a surgical career. Several factors lead us to the conclusion that as the number of female surgeons increase, the number of Full-Time Equivalent's (FTE's) available for patient care will trend lower. The number of female medical graduates has increased from 10% thirty years ago to almost 50% now. Female physicians make up 28.3% of active physicians in the United States in 2007 and the average age for retirement for women physicians was noted to be 4 years earlier then men.[27] In addition, 80% of female physicians are retired by age 65 compared with only 60% of male physicians. Male physicians generally work about 7.4 hours (19%) more patient care hours per week compared with female physicians and since they are over represented in the high stress surgical specialties, their attrition rate is higher compared with female physicians.[28,29]

While supply is deficient, demand for surgical services will likely increase. This is the consequence of the aging of the general population and the greater likelihood that they will require more surgical specialists to care for them. The rate of in patient surgery for people aged 65 and older (4469 per 100,000 people) is 3 times higher than the general population (1519 per 100,000 people).[24] Moreover, from a population based needs projection of the surgical workforce, the fact that women have a longer life span and represent a larger share of the older population makes it likely that women will also require more surgical and specialty care than men.[23]

It is clear that no formula can estimate or predict which of these variables will influence the supply and demand of surgical specialties. It appears based on our analyses that since most of the variables discussed are likely to increase the demand for surgical services and decrease supply, our estimates of the shortage of surgical specialists may indeed be conservative.

Cost Estimates

We also believe we have a reasonable approximation of the future educational costs of training residents in surgical specialties. In Table 3 we noted a shortage of more than 29,000 surgeons and in Table 4 we estimate the total cost of training surgical specialists at the current level at $26.8 billion between 2011 and 2030. The incremental costs for the same time period associated with the additional trainees required to care for the projected population growth are $10.104 billion. These estimates are probably under stated for 2 reasons. First, we are using 2009 dollars as the basis for cost estimates with no adjustment for inflation. Second, we have used Cooper's estimate of $80,000 per trainee per year, which does not include the cost for indirect medical education.

Are Shortages Here Already?

Access to surgical specialists is vital for both the urban and rural population at large. In general, a wait time in excess of 2 to 3 weeks has been used as a sign of a surgical practice needing another associate since waiting time for an appointment indicator is tied to overall patient satisfaction. Consequently, time required to get an appointment with a surgical specialist is a parameter that can be used as an indicator of shortages. A survey in 15 metropolitan areas by Merritt, Hawkins and Associates of new patient wait times gives us an early clue of specialty shortages.[30] For instance, with regard to orthopedic surgery, the reported patient wait times for complaints of joint pain exceeded 14 days in 9 of the 15 (60%) metropolitan markets recently surveyed and over 21 days in 20%. Orthopedic surgery wait times exceeded 21 days in 3 of the 15 metropolitan areas surveyed (20%). Similarly, in OBGYN, average patient wait times were at or exceeded 14 days in 13 of the 15 (87%) metropolitan markets surveyed and 47% were over 21 days.

Another clue about future shortages of surgical specialists is the difficulty recruiting new associates. Already, the number of residents receiving more than 50 job solicitations during their residencies has increased from 16% in 1999 to 52% in 2006.[31] In a recent survey of 398 hospital administrators, 49% described the current situation in recruiting new associates as extremely challenging.[31] This is all the more convincing because 55% of the administrators described the recruitment as having become significantly worse in the last 2 years. The concern for hospital administrators is that the financial impact of a vacancy clearly hurts the hospitals bottom line since for example, the typical time to fill an orthopedic surgery position takes 19 months and results in a net loss for that period to the hospital of over $3.5 million.[31] The predicted shortage of surgical specialties may lead to several undesirable outcomes such as prolonged waiting times for appointments, (Fig. 3) long distance travel for patients in rural areas, pressure on existing specialists to see more patients per hour and burnout, an over reliance on physician extenders and eventual rationing of care.

Figure 3.

Will it come to this?

Small Town and Rural America

Our greatest challenge is the threat to small town and rural America. The shortage in surgical specialists will most directly impact the 54 million Americans who live in rural area and receive care at small hospitals. SOSSUS specifically studied the distribution of surgeons in various size communities and noted that large communities (above 50,000 population) had adequate surgical coverage but rural communities did not.[8] This situation has not changed. Josef Fischer and the Wall Street Journal have pointed out that the shortage of general surgeons is a threat to trauma care both in small towns and in rural America.[32,33] Examination of Table 3 shows the largest single deficit by numbers is in OB/GYN with a gap of almost 14,000, a 27% shortage. Orthopedic surgery has an 18% shortage and GS almost a 10% or more than 2500 surgeons.[32,33] Americans expect to have their babies born and receive trauma and emergency care from these specialists in their local hospitals. With this impending shortage of Obstetricians, Orthopedic Surgeons, and General Surgeons, will this be possible?

Limitations of This Study

Estimation of the physician workforce is an obviously highly complex calculation. One of the problems with using the US Census population forecasts is that Census numbers have been generally proved to be somewhat lower than the actual population growth, which results in high physician to population ratios and an erroneous conclusion of a surplus.[4] In addition, variables such as aging of the population, shortened residency training paradigms, specialization of surgeons to a narrow scope of practice, increased female physicians in the workforce, number of nonphysician clinicians, early retirements of older physicians, the uninsured, and reduced work hours or job sharing have not been taken into consideration. Nor have we addressed the geographic misdistribution of physicians to urban areas rather than small town United States. This analysis also excludes osteopathic trained surgeons who are an important part of the workforce and provide a significant portion of surgical care in rural areas. The state of the economy may also impact the retirement age of surgical specialists. If the retirement accounts stay depleted as they currently are, it may force some surgeons to continue working longer. Finally, disruptive technology could make high volume procedures such as breast biopsies go the way of gastric surgery eliminating an entire class of surgeons. Technology is however, not a 1-way street. Laparoscopic cholecystectomies have more than replaced open procedures and gastric bypasses have supplanted peptic ulcer operations.

Possible Solutions

Increasing the Number of Training Positions Clearly increasing the number of surgical residency positions while eliminating funding caps for post graduate education is necessary. Aggressive steps have been taken to increase the number of graduating medical students. Of the 126 medical schools, 118 have increased their class size.[34] First year MD school enrollment reached 18,036 in 2008 and is expected to be 19,795 by 2013, a 11% and a 20% increase, respectively.[35] But, unless there is a corresponding increase in residency training positions in the surgical specialties, the only outcome will be a reduction in the number of IMG's and not any increase in the number of practicing surgeons in the United States. The number of trainees has gradually increased over the last 5 years fueled by the enforcement of the 80 hours work week rule and the resulting increased need for trainees. The number of spots increased by 7.6% primarily due to IMG's.[36] The number of GME trainees in ACGME programs is 106,012 in 2007 up from 89,368 in 1992 prior to the BBA. The new positions have been funded by hospitals presumably from operating funds. The consequence of a freeze in the funding of GME positions is that the population growth during 1996 to 2006 approached 12.6% with no additional funding being available. The estimated cost of training the surgical workforce to meet the needs of the population is 10 billion dollars over the next 20 years. In an era of trillion dollar deficits, $500 million annually for ten years seems like a modest investment into the future. However, given the significant budget deficits it would seem that the odds of unfreezing the ceiling on GME positions are small in the near future and we have to look for creative, alternative solutions to the problems ahead while continuing to try to convince policymakers of the consequences of ignoring a vast access issue for the public.

To interest trainees in the small town America, existing training programs could receive financial incentives for developing rural surgeon training tracks that would include experiences in small community hospitals. In addition, another proposed option for funding rural residency training is the concept in H.R 2583 The Physician Work Enhancement Act of 2008 passed in the House of Representatives in 2008 and sponsored by Michael Burgess M.D (R-Texas) and Representative Gene Green (D-Texas).[37] This bill to amend title VII of the Public Health Service Act calls for establishing a loan program for eligible hospitals, particularly in rural areas, to establish residency training programs. The loan amounts start at $8 million in 2010 and extend into 2014. Although, this bill only addresses primary care, OBGYN, and mental health physicians, the concept could also be extended to other specialties that experience shortages.

Incentives for Rural Practice The misdistribution of the surgical workforce amplifies the shortages. Incentives could be established to make rural practice more attractive. These may include increased Medicare payment for physicians in these areas and a loan forgiveness program.

Exploration of New Care Models

Patient care models may need to be re-examined and redefined in light of changes in the workforce to create opportunities for part time physicians and encourage job sharing. We must respond positively to the obvious preference of Generation X and Y medical students and younger surgeons who have declared life style issues to be of great importance to their choice of specialty and jobs. The top 5 factors (in decreasing importance) for physicians 35 or younger are: geographical location, call schedule, practice setting, compensation, and professional growth opportunity.[31] Given the large number of surgeons over 55 years of age getting ready for early retirement, promoting job sharing and part-time work to relieve the stress and burnout particularly among female surgeons who are also the primary care givers for their children, seems worth considering. There has been a 50% increase in part-time physicians between 2005 and 2006 most pronounced in senior and midlife physicians.[31] In a recent survey, groups greater than 500 physicians were more likely to have physicians practicing part-time and the overall gender mix of physicians practicing part-time (0.50–0.99 FTE) is dominated by female physicians in a 2 to 1 proportion.[36] The number of female physicians working part-time has also increased from 8% to 12% between 2005 and 2006.[31] Compensation plans for practitioners who wish to work part-time can be custom designed to include a variable overhead structure commensurate with the hours worked. Academic medical centers can also create customized hybrid roles for less than full-time surgical specialists with a choice of a clinical, teaching, or research role. Professional liability premiums must be adjusted lower in proportion to the time of exposure rather than a flat rate. How much impact will postponing retirement and job sharing have on the supply of surgeons? It is estimated that barring a large scale shift, if retirement is postponed for 2 years the difference in physician supply may be in the order of 5%.[23]

Let us use GS as a model for gauging the impact of job sharing on future supply between the years 2010 and 2030. Using our earlier projections of GS, in practice, needed to care for the population in each decade starting in 2010[2] we assumed that 1/3 of retiring GS would work half-time and job share for a total of 5 years thereby providing an additional 117.5 FTE's each year. As a percentage of GS needed, the workforce would increase by 5% in 2010, 5% in 2020, 7% in 2030, and 11% in 2040. While this is but a small part of the overall shortage, the reduced burnout rate and a better lifestyle may convince more medical graduates to enter surgical specialty training.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: