Surgical Education in the United States: Portents for Change

Murray F. Brennan, MD; Haile T. Debas, MD

Annals of Surgery. 2004;240(4) 

The Principle Constituencies of Surgical Education

The United States surgical residency and fellowship education programs have been the envy of much of the rest of the world-structured, focused, monitored, evaluated, and credentialed. Recently, multiple professional, personal, and societal changes have brought pressure to bear on the system, forcing the prospect of major change. Because change is painful, there is a desire to focus on past accomplishments to justify the status quo. However, the winds of influence are such that change is inevitable for the continued supremacy of the postgraduate surgical education system in the United States. We hypothesized that we could define the factors necessitating pressure for change if we looked at the constituency that the education system was designed to serve. We believe that an examination of those expectations would help identify where change is needed. We hoped to sustain a national dialogue on this issue so crucial to our future.

The constituencies being served by the current surgical education and training programs are the patient, the medical student, the surgical resident, the surgical generalist, the surgical specialist, the health care provider/payor, and the hospital/academic department, including the faculty.

Examining each of these to see how well the current education system delivers could, we thought, give insight into the areas requiring the most change and improvement to fulfill future needs.

The Patient

The crucial issue for the patient may well be: Will there be a doctor to care for me?

It is now strongly suggested that by the year 2020, there will be a large shortage in the physician workforce, none more obvious than in the discipline of surgery. If one accepts the postulations based on gross domestic product increases,[1] then by the year 2020, there will be an approximately 200,000 physician shortage in the United States. Although some of this shortfall can be made up by nonphysician clinicians,[2] a serious shortfall in physicians, particularly specialist physicians, is projected.[3]

A recent survey of 70 medical schools in 35 states reported shortages in surgeons that approximate 20%. A survey by the Massachusetts Medical Society suggests a physician shortage in general surgery of 32%, with orthopaedics, neurosurgery, and urology having even greater shortages.

Society's and the patient's perception of what he or she wants has changed. Given a choice between a specialist and a generalist, the patient chooses a specialist.[4] This is compounded by ready access to available information from the Internet, much of it uncensored, colored, and often tainted with institutional, if not personal, claims of supremacy in the delivery of care. The media is vociferous in defining what the public should expect from their surgeons and voracious in its pursuit of perceived inappropriate care. How can we cope with the suggestion that sleep deprivation of residents equates to inebriation?[5]

Some patients look only for a competent generalist. This may be by personal choice of time and access, of distance, or because of a lack of access to a specialist caused by societal, ethnic, financial, or language barriers. Patients roundly rejected the attempt of the managed care industry to create primary care physician gatekeepers for their care.[6] The American public, often well informed about quality of health care, has decided it prefers to be treated within specialties.

Increasingly, patients and employees are asking to see the results of interventions before they select their doctors or hospitals. Clinical care is becoming patient-centered and outcome-based, and it is likely that both referral patterns and reimbursement rates will depend on demonstrated good clinical outcome.[7] Increasingly, patients will have more and more say in their own treatment. As they do, they will no longer accept to be treated by physicians-in-training and will demand that their care be provided by experienced practitioners. This raises the important issue of training by methods that do not involve direct patient care. How, if at all, is our training system prepared to deliver on these important issues?

It seems that in many areas we are perceived as less professional, less caring, and less available. Society is forcing us to give up our profession to be health service delivery workers. This transition has become a fait accompli in some European societies, so are we just fighting the tides of time? It is now certain that we must educate the patient and society that it is not possible to demand limited hours, regulatory oversight, relentless accountability and at the same time competence and unlimited availability. Everyone wants the best care, but no one wants to pay in time, effort, or money.[6]

The Medical Student

Medical school applications by white males as a percentage of baccalaureates has fallen from 1977 to 2000, and in the mid-1980s the applicant to acceptance ratio was the lowest in 25 years. During the period from 1980 to 2000, white Caucasian males were replaced by women (Fig. 1) and non-Caucasians (Fig. 2). Historically, women have shunned surgical training programs, and while women are increasingly considering surgery as a professional career, less than 20% are currently enrolled in surgical training programs.

US medical school graduates by sex, 1979-1980 vs 2002-2003. Source: AAMC Data Book 2004.

US medical school graduates by race, 1979-1980 vs 2002-2003. Source: AAMC Data Book 2004.

For medical students, surgery has become an increasingly unappealing area of medical practice. The National Residency Matching Program[8] suggests that from 1992 to 2000, the numbers ranking general surgery as a preferred discipline fell from 1950 to 1500, an overall decrease of approximately 23%. For US seniors, the fall has been more dramatic, on the order of 36%. The only increase has been in US citizens educated in international medical schools. This number reached a maximum of 140 individuals as of 2002 and, although its overall impact is limited, the impact in terms of the quality of applicants for surgical programs is real. An increase in international medical graduates can compensate for this shortage, but it has inherent problems in terms of the difficulty in evaluating foreign graduate medical education and increasing immigration service review of foreign access to US surgical programs. Active programs to improve international medical graduate performance are currently under way.[9]

Students shun the rigors of surgical training and view the profession as too demanding and providing little opportunity for personal time. From 1996 to 2002, controllable lifestyle explained 55% of variability in specialty preference, controlling for income, work hours, and years of graduate medical education.[10,11] Medical student choices (Fig. 3) have changed to user-friendly specialties.[12] Currently, the median debt at completion of medical school is over $100,000 (Fig. 4). Over 25% of medical school graduates of 2002 incurred a debt greater than $150,000.[13] The prolonged length of surgical training, fellowship, and research experience, requiring 9 to 11 years post-MD at subsistence pay, is poorly perceived, and a strong belief that years are wasted in service without education is a strong disincentive.

Medical school career choices, ranking specialties, 1996 vs 2002. Adapted with permission.[10]

Average indebtedness of medical students. Source: AAMC Data Book 2004.

A powerful factor in the choice of medical specialty is identification with role models. Role models are the impetus to entry into a surgical subspecialty in 56% of residents.[14] What makes an ideal role model has been debated and defined.[15] Role models can be introduced at any point in a training program, perhaps best in medical school. How well do we as surgeons match up to the important attributes that emphasize, as in internal medicine, dedicated time as a teacher, doctor-patient relationships, the teaching of psychological aspects of care, and prior chief resident experience?[15]

A real opportunity exists to provide exposure of medical students to individual electives and mentors at busy clinical services by allowing them to act as assistants rather than students. Such manpower, if allowed, could fill in for losses incurred by workweek hourly restraints and begin the education process sooner, allowing for a decrease in the length of the residency. Such approaches of pre-internship have been used formally in other countries and are widely practiced informally in the United States.

The Resident

Whether surgeons of our vintage appreciate it or not, an inescapable change has taken place in today's residents with respect to their perceived needs for a balanced lifestyle that allows more time with families and free time to follow their interests. We dismiss the lifestyle needs of residents at the peril of our recruitment success.

Currently, clinical service demands take up a large portion of the training program and provide little emphasis on education. The resident perceives the length of training to be too long given the demanding lifestyle and inherent debt at the commencement of training. The advocacy of an 80-hour workweek will have an inevitable impact on the type of candidate exposed. The prospect of 5 to 9 years of residency at $50,000 per year for an 80-hour workweek is certainly insufficient to repay medical school debt, commence a family, or buy a house. Encouragingly, the decline in positions filled in general surgery by US graduates reached a nadir in 2002 at 75.2%, but it was increased to 82.6% in 2003. This was associated with an increase in the percentage of US seniors applying to general surgery from 5.8% in 2002 to 6.5% in 2003. This has been paralleled (Fig. 5) by a decline in the foreign graduate numbers filling PGY1 positions (Fig. 6). The overall implication is that the inherent sacrifices in lifestyle and income during residency without subsequent anticipated financial rewards at completion work counter to the choice of a surgical career.

General surgery positions offered/filled by US graduates in NRMP. Source: AAMC Data Book 2004.

First-year positions in general surgery, declining foreign graduates in PGY1 positions. Source: NRMP Data Book 2003.

Over the years, residents have accumulated more and more administrative chores, many of which make little sense in an environment that is so rich in information systems. Processes by which patients are admitted, cared for, and discharged are antiquated, take minimal advantage of information technology, and impose an unjustifiable burden on residents. These processes need to be totally revised and adapted to current needs. Surgical education has to be more about understanding, imagination, and communication as well as training and skill acquisition. Surgical training programs are perceived as being none of these, but rather focused on service delivery, particularly to the poor and underprivileged.[16] At present, 20% of surgical residents taking the American Board of Surgeons Qualifying examination are female, whereas females constitute 50% of the medical school graduates. The overall contribution of women to the long-term workforce in surgery is limited appropriately by the desire and need for a lifestyle that can accommodate childbearing and childrearing without limiting professional satisfaction. Few residencies have innovative programs to address the important need of women residents and that of male-resident fathers whose spouses have a full-time professional career. The desire for women to enter surgical subspecialties is variable, with increases in female orthopaedic residents lagging behind increases in other areas.[17] We need to recognize that many medical students approach residency together as a couple, requiring adaptability in the residency match (Fig. 7). How well can we adapt to this 2-person professional medical family?

Number of couples applying versus number of couples with both matching for all specialties. Source: NRMP Data Book 2004.

All of these issues compound the need for change. A real opportunity exists to integrate medical students into this educational process, while at the same time decreasing the busywork demands placed on the resident. The experiment in the United Kingdom of double-teaming two women in a training program to allow greater free time for childrearing needs to be examined.

To compound this perception of too little time to learn too much, we impose volume requirements for each resident that are difficult, if not impossible, to meet. The extensive clinical spectra to be covered by any one resident make any proposed decrease in the length of training difficult. We have been reluctant to acknowledge that most institutions and residencies cannot provide the volume and breadth of clinical material that we demand they be exposed to!

The resident facing his board examinations may be justified in complaining when he or she is examined in technical aspects of pancreatic, esophageal, or infected aorto-iliac graft surgery that he or she has rarely seen and never performed.

The General Surgeon

It was suggested as recently as 1998 that most leading surgeons in the country still believe that broad surgical training is superior and should be maintained.[18] It is not surprising that a group (79%) of surgeons older than 50 years would support the regimen that molded them!

The general surgeon, the hallowed product of surgical training in this country, is a vanishing breed. More than 70% of surgeons completing general surgical residency opt for subspecialty training immediately after residency,[19-21] and an even greater majority self-differentiate by the time of first recertification.

Those surgeons who remain as classic general surgeons have a practice that barely resembles what they were trained to do. A recent analysis of the workforce patterns of rural surgeons in West Virginia[21] suggested that more than half were so discouraged that they would not encourage a young person to pursue a career in medicine. More than one third of these surgeons practiced some general medicine daily, and the surgical caseload varied by community size. In communities of fewer than 10,000, they listed obstetrics and gynecologic (9%), urological (5%), otolaryngology (9%), and orthopaedic (4%) procedures as part of the regular caseload. Endoscopic procedures comprised 17% to 24% of total procedures regardless of community size, which is clearly not a focus of the current general residency education program.

Other general surgeons believe that they should be trained and encouraged in all surgical practices, particularly gastrointestinal surgery. Academic reports and societal demands, however, emphasize that the casual operator has morbidity and mortality rates unacceptable for major procedures such as esophagectomy, pancreatectomy, and liver resection.[22,23] High-volume surgeon and institutional experience improves operative mortality, morbidity, and long-term survival.[24] That 25% of patients undergoing pancreatectomy are operated on by surgeons who perform less than 1 procedure a year, and that 96% of surgeons who performed a pancreatectomy in the State of New York between the years 1983 and 1991 performed 1 or fewer annually, is hardly desirable patient care.[23]

Is it possible or even appropriate to train generalists to do such procedures when so few exist in the majority of residency training programs in this country? The converse argument is that while we wish that complex procedures be performed in high-volume centers, it will just not happen; eg, in a study of Medicare patients undergoing pancreaticoduodenectomy[23] 28% of patients had their procedure performed at an institution that did less than 1 a year and 80% had an operation in an institution that did less than 5 a year. However, because it is so does not mean it is the desired aim. Ask the patient! It is becoming clear that when patients have information and the possibility to choose, they will choose centers with better outcomes. Analyzing this issue from the pure perspective of where the operation has been performed is flawed, because it incorrectly assumes that patients had both information and the possibility to choose.

The generalist is in part damned because he or she is judged against the outcome results of the specialist.

More frightening is the question of where might these procedures be performed in sufficient volume to train anyone? In a previous study,[23] only 27 hospitals in the nation did more than 16 pancreatic resections a year, which strongly argues for using these uncommon cases to train only those who will fully embrace this type of practice.[25]

The general surgeon, even in the absence of any fellowship training, self-differentiates. Of surgeons presenting at 10 years for recertification, the average number of hepatic and pancreatic procedures per year was less than 1,[26] 86% did no liver procedures, and 79% had not performed a pancreatectomy. Procedures for which the median annual experience was zero included esophagus, liver, any transplant, splenectomy, and any complex vascular procedure. Even if we believe that the generalist should do it all, the general surgeon and the patient have decided otherwise.

It is clear that we cannot mandate where common operations be performed.[27] It is estimated that if colectomy was limited to hospitals doing a minimum of 15 procedures in 5 years, 1400 patients (5% of the total) would be redirected and 263 hospitals (45%) would discontinue the procedure, and the operative mortality would fall from 4.59% to 4.44%.[28]

What are the recertifying general surgeons doing? In part, it depends on the population, but in the main, it is endoscopy, breast, cholecystectomy, hernias, and appendectomy.

The need for general surgical leadership in clinical practice is very real. It is clear that the general surgeon, whether practicing in small communities or large, remains an important bastion in emergency management of inflammatory conditions such as acute cholecystitis or acute diverticulitis and in initial management of the burned or otherwise injured patient. All of these areas progressively demand the resources of specialized centers and specialized surgeons so that patient outcome can be maximized.

Results of recertifying exams suggest that surgeons, initially general surgeons, progressively differentiate as their practice matures and their identity becomes established. The fate of the generalist is at best unsure and varies between the urban setting and small communities that require comprehensive surgical and medical care. The general surgeon working in urban communities in large nonteaching institutions progressively becomes identified with areas that he or she may or may not have specialist training in, but in which experience substitutes for formal training, and the need or desire to perform the casual operation is no longer necessary.

The Specialist Surgeon

Seventy percent of male surgeons and 50% of female surgeons completing a surgical residency go on to subspecialty training. Many argue that the general surgical experience provides little beyond a basic understanding of surgical principles that prepares them for a subsequent career in a focused discipline. Urology, orthopedics, and neurosurgery take a 1-year internship and then move directly to their own discipline. Would it not be better if we served the needs of all potential surgical faculty with a structured introductory surgical component and then allowed early differentiation?

The primary time when a trainee chooses his or her specialty is as a junior resident[14] influenced, as are medical students, by role models.

The demand for greater cultural training in communication and humanistic as well as technical skills is increasing, but does the dedicated cardiothoracic surgeon need to do one pancreatic resection or total gastrectomy to prepare him or her for a career in which he or she will rarely, if ever, see the organ, let alone operate on it? Many surgical residents who go on to specialty training see at least 2 of the 5 years spent in general surgical training as a waste of the most productive years of their lives, which could be dedicated to their specialty or to research within that specialty. Plastic surgery has experimented with 1, 3, or 5 years of a general surgical program before moving on to a dedicated 2- to 3-year course in plastic and reconstructive surgery. An analysis of the participants in each of these pathways would provide great insight into the merits of each approach.

These problems have been highlighted by the desire of the vascular surgeons to better identify candidates who will pursue vascular surgery and to allow earlier differentiation into that subspecialty. Conversely, the general surgeon who wishes to do some vascular surgery will continue to do so if specialist care is not readily available.

Advances in science and technology are forcing the melding of specialist practices. The training required in vascular surgery needs to be meshed with that of interventional radiology, and combination residencies have been promulgated. A similar dilemma will exist in other image-guided therapies. Conversely, if one wishes to practice general surgery in addition to the specialty field, as has been suggested by one third of current vascular surgeons, how do we address the need for the vascular surgeon to be appropriately and adequately trained for the general surgery practice that he or she desires to pursue?

In a global context, perhaps the greatest challenge is the concept that clinical care is becoming disease focused rather than discipline focused. One will enter a cardiovascular, cancer, or burn institute or a trauma program where the focus is the disease rather than the discipline. Complete familiarity with what is available in other disciplines will be essential to the wise and appropriate practice of the surgical subspecialist. Advances made in genetic identification of risk groups, the need for genetic and molecular diagnosis, and the use of genetic therapy reinforces this approach.

The demands for analysis of competency, whether in training, generalist, or specialist practice, will add further burdens to the limits to which comprehensive all things to all people training can be maintained. How appropriate is it to demand of a vascular surgeon that he or she be recertified in general surgery, or that a general surgeon be recertified to include the elements of vascular surgery that he or she no longer practices? What of the person who has dedicated his or her life to the management of breast disease, who is asked to be recertified in the broad discipline that he or she initially embarked upon? Testing surgeons in areas in which they do not practice defeats the goal of recertification or continuing education. How much does recertification in a comprehensive generalist sense, when personal focus is highly specialized, deter re-evaluation and assessment of competency?

The Payor

And what of the ultimate constituency-the payor? Whether the payor is the federal government, the insurance carrier, or the state, all look for value for money. State registries can identify and characterize the practice patterns of individual surgeons, the mortality, morbidity, length of stay, and other characteristics that track or evaluate some degree of competency of care. Increasingly, data is available for surgeon, by operation, for the cost and charges initiated by that provider. This is compounded by cost analysis and increasing demands to lower the cost of medical care when the demand for expensive technology is simultaneously increasing. Demand for technologically advanced clinical care is increasing as societal exposure and media promises escalate. The nexus of ever-increasing technology with ever-increasing demands by society for its use creates the impossible prospect of limitless expenditure, which is a prospect incompatible with the trends in the national economy. Why are we so unwilling to add cost-effectiveness to the evaluation of the clinical efficacy of each newly introduced technology? The evaluation of any new clinical or technical approach should be by the physician and patient who use it, not by the discipline or industry that invented it! Only if leadership is provided in the application of technology, with an appropriate evaluation of benefit, can we possibly limit runaway application.

Surgeons, on occasion, have taken extreme responses to pressures of legislation and registration. Look to the current federal and state debate on malpractice and the limits on malpractice awards. Consider the decision by some to opt out of the Medicare system providing boutique practices or indeed enrolling patients in personal managed programs of health care delivery. All these pressures show little sign of ameliorating. The constituencies of payor and surgeon, both seen as losing, is an area of great dissatisfaction. A better solution than just opting out is needed.

The Hospital and Surgical Departments and the Faculty

The hospital and the academic surgical department, in which surgical training occurs, is the constituency in which the demands for change must be implemented. In the past 50 years, surgical residents have provided services critical to hospital operations. Much of the patient care services residents provide are appropriate to surgical training, and Medicare compensates hospitals for the service the residents provide, but some of the services residents provide have no intrinsic educational value. As hospitals become increasingly challenged financially by declining reimbursement, they eliminate hospital services that residents and nurses then have to absorb. With the 80-hour workweek regulations now in effect, hospitals will have to recreate these services. As a result, how graduate medical education funds are used will come under greater scrutiny.

Teaching can impose a degree of inefficiency in the provision of hospital services, and typically the cost structure of teaching hospitals is 25% to 30% higher than that of community hospitals. Thus, hospitals will scrutinize and demand appropriate reimbursement for the cost of this necessary inefficiency if they are to survive in an arena of brutal health market competition.

Faculty in medical schools already demonstrate a 30% level of high emotional exhaustion and burnout, with younger surgeons more affected![29] Given the current faculty workload of between 60 and 80 hours per week (Fig. 8), it is unlikely they can absorb more teaching and clinical cases.[30] Exhausted faculty are poor role models!

Average total hours worked per week by practice type. Reprinted with permission.[30]

Residency programs are the pride of academic surgical departments and provide the single most important focus for faculty unity. As reimbursements for physician services decline and faculty work longer hours and cope with increasing documentation, their ability to devote adequate time to teaching and other scholarly activities has been strained. Most departments do not make specific financial commitments for teaching. It is fortunate that faculty members receive gratification from teaching and their association with residents in the care of sick patients. The lack of specific compensation for teaching will in the long term degrade the learning environment and challenge the traditional ad hoc understanding under which surgical education is provided. How and by whom should surgeons be financially compensated for the teaching they do? How will we accept that the process of education is an equally fruitful source of research and evaluation? A desirable pathway to surgical fulfillment is currently poorly recognized and inadequately compensated.

How will training programs square the competitive demands of a decreasing workforce, decreased work hours, shortened residencies, and lack of surgical volume in specialist areas with residents' need to fulfill volume requirements in a broad spectrum of disease areas and technical skill acquisition? How can we franchise the junior faculty, the bastion upon which any future success is dependent?


We have attempted to examine the changes that are affecting the key constituencies of surgical education. In so doing, we have uncovered many areas for which the traditional education and training system falls short of fulfilling expectations. We have identified some of the profound changes that have affected patients and society at large, requiring a significant change in the way we educate the next generation of surgeons. Failure to respond in a thoughtful, strategic, and prospective manner to the multitude of challenges faced by key constituencies can only result in further decreases in the attractiveness of a professional lifestyle that was once envied and of extraordinary reward. The experiences of other countries are that professionalism has been lost and the surgical technician becomes another member of the nonprofessional health care delivery workforce. This change is one that many of our generation would like to see forestalled or avoided. Only by addressing the concerns of all constituents about the product we are delivering can the Untied States maintain its hard-won leadership role in surgical education. We believe that the leadership of the surgical discipline in this country is best suited to develop and implement the necessary changes. This challenge presents a moral obligation of gargantuan proportions to this generation of surgeons and at the same time a unique opportunity to shape the future of our profession and the way we serve our patients. Change takes time, but it begins with acceptance of the need for change. We need to educate in what we practice and certify by what we do.


  1. Cooper RA. Medical schools and their applicants: an analysis. Health Affairs. 2003;22:71-84.

  2. Cooper RA, Stoflet SJ. Diversity and consistency: the challenge of maintaining quality in multidisciplinary workforce. J Health Serv Res Policy. 2004;9:39-47.

  3. Sheldon GF. Great expectations: the 21st century health workforce. Am J Surg. 2003;185:35-41.

  4. Fischer JE. Our health care system: where are we going? Bull Am Coll Surg. 2000;3:25-42.

  5. Dawson D, Reid K. Fatigue, alcohol and performance impairment. Nature. 1997;388:235-237.

  6. Thompson JC. Seed corn: impact of managed care on medical education and research. Ann Surg. 1996;223:453-463.

  7. Birkmeyer JD, Finlayson EV, Birkmeyer CM. Volume standards for high-risk surgical procedures: potential benefits of the Leapfrog initiative. Surgery. 2001;130:415-422.

  8. National Residency Matching Program. Available at:

  9. Horvath K, Coluccio G, Foy H, Pelegrini C. A program for successful integration of international medical graduates into US residency training. Curr Surg. 2004; in press.

  10. Dorsey ER, Jarjoura D, Rutecki GW. Influence of controllable lifestyle on recent trends in specialty choice by US medical students. JAMA. 2003;290:1173-1178.

  11. Morris JB, Leibrandt TJ, Rhodes RS. Voluntary changes in surgery career paths: a survey of the program directors in surgery. J Am Coll Surg. 2003;196:611-616.

  12. Lind DS, Cedan JC. Two decades of student career choice at the University of Florida: increasingly a lifestyle decision. Am Surg. 2003;69:53-55.

  13. Robinson L, ed. AAMC Data Book: Statistical Information Related to Medical School and Teaching Hospitals. Washington, DC: Association of American Medical Colleges; 2004.

  14. Ko CY, Whang EE, Karamanoukian R, et al. What is the best method of surgical training? A report of America's leading senior surgeons. Arch Surg. 1998;133:900-905.

  15. Wright SM, Kern DE, Kolodner K, et al. Attributes of excellent attending physician role models. N Engl J Med. 1998;339:1986-1993.

  16. Bolton G. Medicine, the arts, and the humanities. Lancet. 2003;362:93-94.

  17. Blakemore LC, Hall JM, Biermann JS. Women in surgical residency training programs. J Bone Joint Surg. 2003;85(-A):2477-2480.

  18. Bland K. Challenges to academic surgery: the impact of surgical fellowships on choice of an academic career. Bull Am Coll Surg. 2000;85:17-24.

  19. Kwakwa F, Jonasson O. The longitudinal study of surgical residents 1994-1996. J Am Coll Surg. 1999;188:575-585.

  20. Rattner DW, Apelgren KN, Eubanks WS. The need for training opportunities in advanced laparoscopic surgery: the residents' perspective. Surg Endosc. 2001;15:1066-1070.

  21. Gates RL, Walker JT, Denning DA. Workforce patterns of rural surgeons in West Virginia. Am Surgeon. 2003;69:367-371.

  22. Begg CB, Cramer L, Hoskins WJ, et al. Impact of hospital volume on operative mortality for major cancer surgery. JAMA. 1998;280:1747-1751.

  23. Lieberman MD, Kilburn H, Lindsey M, et al. Relation of perioperative deaths to hospital volume among patients undergoing pancreatic resection for malignancy. Ann Surg. 1995;222:638-645.

  24. Skinner KA, Helsper JT, Deapen D, et al. Breast cancer: do specialists make a difference? Ann Surg Oncol. 2003;10:606-615.

  25. Ritchie WP Jr, Rhodes RS, Biester TW. Work loads and practice patterns of general surgeons in the United States, 1995-1997: a report from the American Board of Surgery. Ann Surg. 1999;230:533-543.

  26. Birkmeyer JD, Warshaw AL, Finlayson SRG, et al. Relationship between hospital volume and late survival after pancreaticoduodenectomy. Surgery. 1999;126:178-183.

  27. Birkmeyer JD, Siewers AE, Finlayson EVA, et al. Hospital volume and surgical mortality in the United States. N Engl J Med. 2002;346:1128-1137.

  28. Schrag D, Cramer LD, Bach PB, et al. Influence of hospital procedure volume on outcomes following surgery for colon cancer. JAMA. 2000;284:3028-3035.

  29. Campbell DA Jr, Sonnad SS, Eckhauser FE, et al. Burnout among American surgeons. Surgery. 2001;130:696-702.

  30. Winslow ER, Bowman MC, Klingensmith ME. Surgeon workhours in the era of limited resident workhours. J Am Coll Surg. 2004;198:111-117.