General Surgeon Model Components

Published 2019

In this module:

This module describes adaptation of HWSM to model supply and demand for general surgeons. This work is in response to Senate Report 115-289:1

Congress urges HRSA, to study access by underserved populations to general surgeons and provide a report to the Committee 18 months after enactment detailing potential surgical shortages, especially as it relates to geographic location (i.e., rural, urban, and suburban). For the report to the Committee, HHS should consult with relevant stakeholders, including medical societies, organizations representing surgical facilities, organizations with expertise in general surgery, and organizations representing patients.

The research team reached out to key stakeholders to identify the best available data sources, discuss trends affecting general surgery workforce supply and demand, and provide the opportunity for feedback on preliminary findings. The research team met with the American College of Surgeons to share preliminary findings and solicit feedback. The information provided in this technical documentation and in HRSA reports does not necessarily reflect the views of the organizations that responded to our invitation to participate in the workforce study. There may not be clear consensus on all assumptions. Other organizations representing surgeons, surgical facilities, and patients were contacted, but these organizations did not express interest in participating.

Modeling Supply 

Estimating the Base Year Workforce Supply 

Estimates of the starting supply of general surgeons come from the 2017 American Medical Association (AMA) Masterfile. Key variables include physician age, sex, activity status, and primary work location (which is mapped to state and county). The Masterfile contains information on physicians’ first and second specialties. The following criteria were used to categorize physicians as general surgeons based on this information. 

  • Physicians were first categorized as surgeons if either their first or second specialty is for surgery. (Note that some physicians in the Masterfile might have a record of first or second specialty being a non-surgical specialty with the other specialty listed being a surgical specialty).  
  • Surgeons were categorized as a specialist if either their first or second specialty listed is a surgical sub-specialty. Otherwise, they were categorized as a general surgeon. 

A total of 24,391 active physicians were identified as general surgeons in the 2017 AMA Masterfile. Active is defined as working at least 20 hours per week in professional activities. For purposes of modeling, only physicians who have completed their graduate medical education are counted as active in the workforce. Of these active general surgeons, 19,436 (80%) list general surgery as their first specialty with their second specialty as blank. Another 1,692 physicians list general surgery as their second specialty and (a) list a non-surgical specialty as their first specialty or (b) list another surgical specialty that is included under general surgery. The remaining 3,263 physicians list a surgical specialty that is modeled with general surgery, which include: 

  • abdominal surgery 
  • complex general surgery oncology 
  • pediatric surgery 
  • pediatric transplant 
  • surgical critical care 
  • surgical oncology 
  • transplant surgery 
  • traumatic surgery 

The most prevalent non-surgical specialties listed are anesthesiology, emergency medicine, and diagnostic radiology.  

Rurality associated with a general surgeon is dictated by the rurality of the county of his/her office location listed in the AMA data. Counties are assigned one of six designations of rurality by the CDC’s National Center for Health Statistics Urban-Rural Classification Scheme for Counties.2 These 6 classifications were grouped into one of four designations. Counties classified as “Large central metro” are considered urban. Those classified as “Large fringe metro” are considered suburban. Those classified as “Medium metro” or “Small metro” are combined under the category town. Counties classified as “Micropolitan” and “Non-core” are considered rural. 

Modeling New Entrants 

To model additions to the workforce each projection year, a synthetic population is created. New entrants are drawn from this synthetic population for the simulation such that the number of newly created individuals reflects the annual number and demographics of recent new graduates. The estimated number of general surgeons completing their residency training in 2017-2018 is 1,210. This consists of 1,092 general surgeons completing residency from an Accreditation Council for Graduate Medical Education program and 118 from American Osteopathic Association accredited programs.3 The AMA Masterfile contains the year each general surgeon completed his or her graduate medical education. The distribution of new graduates between 2010-2017 by state, age, and sex defines the demographic distribution of new surgeons for the projection simulation. New surgeons were also assigned a rurality level based on the rurality distribution of general surgeons who recently completed training. 

Modeling Workforce Participation 

For general surgeons age 50 and over, the probability of retirement in any projection year is modeled based on analysis of Florida’s 2012-2013 Physician Survey. In this survey, physicians were asked about their intention to retire over the upcoming three years.4 The attrition probabilities are specific to general surgeons and increase with age. The sample size is insufficient to model attrition by physician sex. The projections assume no attrition from the national surgeon labor force for surgeons under age 50. While there is insufficient data to estimate attrition prior to age 50, attrition among surgeons under age 50 likely is very small. 

Modeling Hours Worked 

Hours worked patterns were estimated from regression analysis of survey data for general surgeons across four states (FL, SC, NY, MD) covering the period 2012-2018.5 The dependent variable is hours worked per week in patient care activities. Explanatory variables consist of dichotomous variables for 5-year age groups and sex, as well as interaction terms between age group and sex.6 Supply projections reflect the changing demographics of the workforce and differing average hours worked per week based on surgeons’ age and sex. The expected number of hours worked weekly by each simulated surgeon is converted to FTE supply by dividing the person-hours worked by 40. This creates a uniform standard of 1 FTE as working 40 hours per week. This also means that the baseline FTE generated by HWSM can differ from the actual count of general surgeons employed.  

  • 2Centers for Disease Control and Prevention. 2013 NCHS Urban-Rural Classification Scheme for Counties.
  • 3Brotherton SE, Etzel SI. Graduate Medical Education, 2017-2018. JAMA. 2018;320(10):1051.
  • 4Efforts continue to find improved data sources for modeling physician attrition patterns. One potential future source is a survey of physicians recently conducted by the Association of American Medical Colleges which collects information on physician retirement intentions. Past analyses have compared the age distribution of physicians across subsequent years of the AMA Masterfile, but there were concerns that due to the lag between physicians leaving the workforce and when that information becomes available in the Masterfile that attrition patterns were under-stated.
  • 5State survey data is used for modeling hours worked, rather than the ACS which is used for modeling hours worked for many occupations, because ACS does not identify physician specialty and hours worked patterns vary substantially by specialty.
  • 6Dichotomous variables take on the value of 1 if the person is in that specified group, and 0 otherwise.

Modeling Demand 

This section discusses issues specific to projecting demand for general surgeons. See Modeling Demand for Health Care Services and Providers for a discussion of the general approach to modeling demand for health care services and providers employed by the HWSM.  

The main components of modeling demand for general surgeons are: 

  • a database containing a representative sample of the population in each county including values for each resident’s characteristics used to predict demand for health care services 
  • prediction equations (based on patient characteristics) for both annual number of ambulatory visits to general surgeons and use of hospital-based care (emergency visits and hospitalizations 
  • staffing patterns that translate demand for services to demand for FTE general surgeons.  

Geographic Variation in General Surgeon Scope of Work 

Procedures performed by general surgeons overlap procedures performed by specialist surgeons. Rural areas tend to have fewer specialists than urban areas. In these rural areas, the mix of procedures performed by general surgeons differs from their urban counterparts (data analytics described below). Efforts to improve access to surgery in rural areas are more likely to focus on attracting and retaining general surgeons than on attracting and retaining specialist surgeons.7 As such, HWSM’s general surgeon projections model geographic differences in practice patterns, rather than apply national patterns to rural areas. We also assume that higher demand for general surgeons in rural areas is at least partially due to decreased availability of specialist surgeons. 

To quantify urban-rural differences in general surgeon practice patterns, the number of procedures performed by physician specialty across three levels of rurality (super rural, rural, and urban8 ) were tallied from 2017 Medicare data. Since surgical procedures differ by the amount of time surgeons spend providing services, work relative value units (WRVUs) were used as a proxy for physician time. Each type of services is weighted by WRVU. 

There were 100 procedure categories in the CMS data where work is provided in 2017 by general surgeons to Medicare patients. Of these, 15 categories account for 93% of total WRVUs (Exhibit 31). By far the largest category in terms of WRVUs is “Major Procedure – Other.” 

For procedures that are categorized by rurality, among all procedures performed by general surgeons for Medicare patients in 2017, one fourth (24.5%) of total WRVUs were provided in super rural or rural areas. Another three fourths (75.5%) were performed in urban areas (Exhibit 32). The distribution across rurality designations differed somewhat by procedure category. For example, for procedure category “Endoscopy - Upper G.I.” in urban areas general surgeons accounted for only 65.2% of WRVUs. 

Analyzing the above 15 procedure categories across all surgeon specialties, general surgeons practicing in urban areas generated 23% of WRVUs in urban areas. General surgeons practicing in rural areas generated 26% of WRVUs in rural areas (Exhibit 33). These urban/rural proportions differ by procedure category. For the category “Endoscopy - Upper G.I.”, general surgeons generated 55% of WRVUs in urban areas and 69% of WRVUs in rural areas. This appears to be driven largely by the availability of thoracic surgeons in urban areas to perform such procedures.  

Practice patterns of general surgeons in rural versus urban areas suggest that per capita use of general surgeons to provide care to Medicare beneficiaries is 13.1% higher in rural areas compared to urban areas. This accounts for the lower availability of specialist surgeons in rural areas. Based on this, HWSM assumes overall demand for general surgeons is 13.1% higher in rural areas than urban areas regardless of insurance type of patient. 

To estimate starting year (2017) demand, the national shortfall of general surgeons for purposes of calibrating the demand portion of HWSM, starts with the assumption that supply and demand for general surgeons are roughly in equilibrium nationwide, though with substantial geographic maldistribution of supply, but that in rural areas not only is there a shortfall from maldistribution but also additional demands on general surgeons due to a lack of specialist surgeons. Based on analysis of Medicare claims data, we calculate that the per capita demand for general surgeons in rural areas by is 13.1% higher than the level that would be required if general surgeons had similar practice patterns to their metropolitan-based counterparts. Applying this 13.1% increase to general surgeon demand in rural areas equates to a 660 (2%) national shortfall. Hence, national estimated demand for general surgeons in 2017 is estimated at 29,980 FTEs (29,320 FTE supply plus 660 FTE shortfall). The estimated shortfall in rural areas is much greater than this 660 national shortfall estimate reflecting that the shortfall in rural areas consists of both the shortfall associated with geographic maldistribution plus the additional 660 general surgeons required in rural areas to compensate for the paucity of specialist surgeons (or equivalent to a 2,490 total shortfall in rural areas in 2017, suggesting that supply of 3,250 is only sufficient to provide 57% of total demand for general surgeon care).

Geographic Variation in Hospital Location 

In this section, we examine demand for care where patients are located. We also examine demand for care where hospitals that employ surgeons or where surgeons might have privileges are located. This distinction is important to interpret HWSM demand results. The population in a rural area will have demand for general surgeons. If, however, there is no hospital in that area then there will not be employment opportunities for general surgeons. Hence, adequate access to general surgery services combines demand within a given area, adequate hospital infrastructure in that area, and sufficient supply of general surgeons to meet demand. 

HWSM models demand for health care services and the derived demand for health care providers based on where patients reside. Employment opportunities for general surgeons (which influences surgeon supply) reflect where hospitals are located. Few hospitals are in rural areas. These hospitals tend to be much smaller than hospitals in more populated areas. Consequently, patients in rural areas often need to travel to hospitals in more populated areas to receive care—particularly complex care. Likewise, within metropolitan areas, larger hospitals and teaching hospitals treating more complex cases are often found closer to urban centers. Hence, some care for people residing in suburban areas is received in urban centers. 

The total number of beds nationwide in hospitals likely to employ general surgeons in 2018 is approximately 858,100.9 These types of hospitals are general acute care, children’s, critical access, and military. The distribution of these beds across counties indicates: 

  • 36% of beds were in large central metro areas 
  • 18% in large central fringe areas 
  • 22% in medium metro areas 
  • 11% in small metro areas 
  • 14% in rural areas 
    • 9% in micropolitan areas 
    • 5% in non-core areas (Exhibit 34

The average bed size of these hospitals ranges from 297 in urban areas to 38 in non-core rural areas. When combining urban centers and the suburbs that surround them, these large metropolitan areas have approximately 53% of the nation’s hospital beds and 58% of general surgeons. About 50% of national demand for general surgeon services comes from the population living in these large metropolitan areas. Combining medium and small metros, the account for approximately 33% of beds, 31% of supply, and 32% of demand. Rural areas have 14% of beds, 11% of general surgeon supply, and 19% of total general surgeon demand. These findings underscore the nuances of interpreting HWSM supply and demand projections for general surgery. Looking at each of the NCHS urban-rural classifications in isolation could lead to misunderstanding of results when considering where hospital infrastructure is located and patient travel patterns to receive some types of care.

Status Quo Demand Scenario 

The Status Quo demand scenario in HWSM assumes the continuation of current patterns in use and delivery of surgery-related services. It projects demand for general surgeons accounting for: 

  • temporal and geographic variation in demographics 
  • prevalence of risk factors 
  • disease prevalence 
  • economic factors correlated with demand for surgery 

There is a small effect of expanded insurance coverage reflecting that several states are in various stages of expanding Medicaid access. The impact on demand for general surgeons from this expanded insurance coverage is small. Most of the insurance expansion effects of ACA already have been realized, and proposed state plans for Medicaid expansion are moderate. This scenario accounts for the limited supply of specialist surgeons in rural areas. It assumes that care use and delivery patterns will remain relatively unchanged over the foreseeable future. 

Evolving Care Delivery Scenario 

The evolving care delivery scenario builds on the Status Quo scenario that captures the growth in demand associated with population growth and aging. In addition, this scenario models other trends and national goals that could affect demand for general surgeons. Specific trends and/or national goals that were reviewed are summarized below: 

  • Achieving select population health goals: This scenario component is described in Section III.D. It involves modeling the health and longevity implications of achieving moderate improvements in national goals. These goals are to reduce excess body weight, smoking cessation, and improved control of blood pressure, cholesterol, and blood glucose levels. Analysis with the Disease Prevention Microsimulation Model suggests that achieving modeled goals could reduce annual demand for general surgeons by 1-2% (up to 610 FTEs) over the subsequent five years due to improved population health. However, by 2030, demand is estimated to be about 3% (1,140 FTEs) higher than the Status Quo projections due to increased longevity and thus a larger and older population still living. 
  • Accountable care organizations: Despite the growth of ACOs, there is little evidence to suggest an impact on demand for general surgeons. An analysis of Medicare inpatient spending for six common elective surgical procedures found no ACO impact on Medicare payments or clinical outcomes for patients receiving surgery.11 There does appear to be a statistically significant 0.52 percentage point decline in hospital readmissions following major surgery associated with ACO participation, equivalent to 4,410 hospitalizations avoided. The implications for demand of general surgeons is unknown, but likely small.12 Other studies found no impact on readmissions following cancer surgery13 and no impact on reducing disparities in rates of surgical intervention by race and ethnicity.14
  • Bundled payments: Literature on bundled payments was reviewed to ascertain whether evidence exists of resulting changes in services utilization or demand for providers. An evaluation of Medicare claims from 2013-2015 found no statistically significant changes in length of stay, emergency department use, or readmission after hospital discharge.15 No published papers on how bundling might affect demand for general surgeons were identified. 
  • Technology: The potential exists for disruptive technologies to transform how surgical care is delivered. Robotically-assisted procedures are increasingly becoming an alternative to traditional laparoscopic or open surgery.16 17 However, there is limited evidence that such surgery will dramatically change demand for general surgeon services.18 19 Like telemedicine, though, robot-assisted surgery could potentially help break down geographic barriers to receiving some types of care.  

There is substantial uncertainty about how surgical care might evolve over time. Efforts to increase population health have counterbalancing effects. One effect is to reduce demand by improving people’s health. Another effect is increased demand by increasing longevity—especially among an older population. Changes in imaging and computer-assisted surgery technology could potentially reduce surgeon operating time thereby reducing demand. These advances could also make possible new types of care that currently is unavailable. For this scenario, the change in demand associated with achieving the predicted levels of the population health outcomes that are goals of the health care system were modeled. Also modeled were the implications of implementing value-based insurance design and care models such as patient centered medical homes and team-based care.20 21 22 23 24 25

  • 11Nathan H, Thumma JR, Ryan AM, Dimick JB. Early Impact of Medicare Accountable Care Organizations on Inpatient Surgical Spending. Ann Surg. 2019;269(2):191-196. doi:10.1097/SLA.0000000000002819.
  • 12Borza T, Oerline MK, Skolarus TA, et al. Association Between Hospital Participation in Medicare Shared Savings Program Accountable Care Organizations and Readmission Following Major Surgery. Ann Surg. 2019;269(5):873-878. doi:10.1097/SLA.0000000000002737.
  • 13Herrel LA, Norton EC, Hawken SR, Ye Z, Hollenbeck BK, Miller DC. Early impact of Medicare accountable care organizations on cancer surgery outcomes. Cancer. 2016;122(17):2739-2746. doi:10.1002/cncr.30111.
  • 14Schoenfeld AJ, Sturgeon DJ, Dimick JB, et al. Disparities in Rates of Surgical Intervention Among Racial and Ethnic Minorities in Medicare Accountable Care Organizations: Annals of Surgery. 2019;269(3):459-464. doi:10.1097/SLA.0000000000002695.
  • 15Joynt Maddox KE, Orav EJ, Zheng J, Epstein AM. Evaluation of Medicare’s Bundled Payments Initiative for Medical Conditions. New England Journal of Medicine. 2018;379(3):260-269. doi:10.1056/NEJMsa1801569.
  • 16Farivar BS, Flannagan M, Leitman IM. General Surgery Residents’ Perception of Robot-Assisted Procedures During Surgical Training. Journal of Surgical Education. 2015;72(2):235-242. doi:10.1016/j.jsurg.2014.09.008.
  • 17Stewart CL, Ituarte PHG, Melstrom KA, et al. Robotic surgery trends in general surgical oncology from the National Inpatient Sample. Surg Endosc. 2019;33(8):2591-2601. doi:10.1007/s00464-018-6554-9. 
  • 18Yaxley JW, Coughlin GD, Chambers SK, et al. Robot-assisted laparoscopic prostatectomy versus open radical retropubic prostatectomy: early outcomes from a randomised controlled phase 3 study. The Lancet. 2016;388(10049):1057-1066. doi:10.1016/S0140-6736(16)30592-X.
  • 19Haese A, Knipper S, Isbarn H, et al. A comparative study of robot-assisted and open radical prostatectomy in 10 790 men treated by highly trained surgeons for both procedures. BJU Int. 2019;123(6):1031-1040. doi:10.1111/bju.14760. 
  • 20Agarwal R, Gupta A, Fendrick AM. Value-Based Insurance Design Improves Medication Adherence Without An Increase In Total Health Care Spending. Health Affairs. 2018;37(7):1057-1064.
  • 21Collinsworth A, Vulimiri M, Snead C, Walton J. Community Health Workers in Primary Care Practice: Redesigning Health Care Delivery Systems to Extend and Improve Diabetes Care in Underserved Populations. Health Promotion Practice. 2014;15(2_suppl):51S-61S. doi:10.1177/1524839914539961.
  • 22DeVore S, Champion RW. Driving Population Health Through Accountable Care Organizations. Health Affairs. 2011;30(1):41-50.
  • 23Farley JF, Wansink D, Lindquist JH, Parker JC, Maciejewski ML. Medication adherence changes following value-based insurance design. Am J Manag Care. 2012;18(5):265-274.
  • 24Green LA, Chang H-C, Markovitz AR, Paustian ML. The Reduction in ED and Hospital Admissions in Medical Home Practices Is Specific to Primary Care–Sensitive Chronic Conditions. Health services research. 2018;53(2):1163-1179.
  • 25Hibbard JH, Greene J, Sacks RM, Overton V, Parrotta C. Improving population health management strategies: identifying patients who are more likely to be users of avoidable costly care and those more likely to develop a new chronic disease. Health Services Research. 2017;52(4):1297-1309.
Date Last Reviewed: