In 2002, an ASA-commissioned survey found that 47% of hospital administrators said their hospital “had to limit the number of operating rooms in service or the available hours of an operating room due to a lack of availability of anesthesia providers”; 59% said “the anesthesia group at their hospital was actively recruiting.”7,8 Of large hospital (>250 beds) administrators, 75% reported an “increase in patients' surgery wait time based on availability of anesthesia care,” 66% limited opening available ORs because of anesthesia personnel availability, and 47% said that, “in their opinion, their hospital did not have an adequate number of anesthesiologists on staff.”7,8 In the 2003 Society of Academic Anesthesiology Chairs/Association of Anesthesiology Program Directors (SAAC/AAPDb) survey, 26% and 15% of responding academic departments reported that they “curtailed service by reducing or closing anesthetizing locations due to lack of faculty or certified registered nurse anesthetists” (CRNAs), respectively. Responding to a 2003 survey, 44% of 114 medical school deans and executives of state professional societies noted shortages of physicians in the specialty of anesthesiology, and 89% reported shortages in at least 1 medical specialty.9 The specialties having the most severe shortages were anesthesiology, radiology, and primary care.
This ASA-commissioned survey completed by the Tarrance Group did not include freestanding ambulatory surgery facilities, office-based practice, or anesthesiology pain practices. Based on the results of this survey and, assuming a vacancy rate per hospital of 1.6, shortage estimates ranged between 1600 and 2200 vacancies for all larger hospitals. Further assuming that an average of 3 rural hospitals are served by 1 group, added at least 100 vacancies, yielding 1700 to 2300 open positions in hospitals alone. These considerations did not include service growth in ambulatory surgery facilities, office-based practice, or intensive care and pain practices. Readers should be aware that the Tarrance Group survey results were not published in the peer review literature.
Further evidence came from position vacancy and salary data, because salary increases could be considered a delayed marker of demand for anesthesia personnel. Medical school deans indicated that a “disruption” in salary structures was taking place, with a constant threat of losing physicians to community practices.9 Indeed, evidence from the Medical Group Management Association (MGMA) shows a 13.4% increase in salaries nationwide with regional increases as high as 32.7% in states such as New York during 2000 to 2003 (Table 2a). Likewise, New York's need for pain physicians grew; job offers increased from 4.4 in 2000 to 6.2 in 2003.10 In 2003, 78% of academic anesthesia departments had open faculty and 64% had vacant CRNA positions.
Survey respondents reported 4 open faculty positions per department, up from 3 in 2002 but unchanged from 2000 and 2001.11 Extrapolating to all academic anesthesiology programs, one would have estimated that there would be 568 open positions in 2003; if 20% of the anesthesiologist workforce is assumed to be in academic departments and assuming that academic vacancy rates approximated those in private practice, a national shortage estimate based on the survey would have amounted to approximately 2800 anesthesiologists. At a time when the anesthesia workforce shortage was still substantial, academic anesthesia departments were confronted with the adoption of resident work-hour restrictions. In July 2003, the Accreditation Council for Graduate Medical Education (ACGME) adopted a stringent set of work-hour rules for all residency programs.12 The impact of ACGME's work-hour rules on academic anesthesia departments was evaluated in the 2003 SAAC Practice Survey. To make up for clinical coverage losses from Residency Review Committee (RRC)-mandated residency work-hour restrictions, 2 added residents, 3.5 CRNAs, and 2.4 anesthesiologist faculty members were needed. Position vacancies and compensation trends confirmed the existence of a continuing anesthesia personnel shortage during 2000 to 2003, representing a deficit of 2000 to 3800 anesthesiologists in 2002 to 2003 (Table 2b).
A survey of >4000 medical/surgical hospitals and ambulatory surgery centers (ASCs) was conducted by the American Association of Nurse Anesthetists (AANA) in late 2002, collecting responses from >1000 hospitals and 600 ASCs.13 Survey results estimated a national workforce of 28,000 CRNAs, and indicated that supply of new CRNAs had increased but such increases had not been fast enough to offset the number of retiring CRNAs. The average age of the CRNA workforce was 47 years and was expected to increase through the year 2018. Fifty percent of CRNAs with PhD degrees were expected to retire by 2012, potentially limiting the educators needed to train larger numbers of new CRNAs. For the year 2002, 3277 US vacancies for CRNAs were estimated, with 2822 in hospitals and 455 in ASCs. Ohio CRNA vacancies were estimated at 134, which is higher than the number of vacancies advertised in www.gaswork.com or the number extrapolated from the 2003 Ohio Society of Anesthesiologists' membership survey (Table 2c).
Initial Recovery from the Anesthesia Workforce Shortage: 2004 to 2007
During these years, the anesthesiologist workforce grew approximately 2% annually (Table 2b), entry of US medical students increased substantially and that of foreign-trained physicians decreased.14 Graduates of US medical schools entered anesthesiology at a high, yet stable rate during 2004 to 2007. As seen from Figure 2, National Residency Matching Program (NRMP) recruitment reached an all-time high, with match positions having increased steadily, from 946 in 1996 to 1338 in 2007. Yet the total number of cohort anesthesiology residency positions, as approximated by the number of graduating residents, increased more slowly to approximately 1500 by 2007. During these years, the number of graduating residents stayed relatively constant because of federally mandated funding caps and continuing attrition on the order of approximately 80 residents per 3-year cohort (Fig. 2).
Although in 2004 only 1200 match positions were filled, by 2007, this number had grown to 1286, an increase in match recruitment of 7.2% during this 3-year period. At a then historical peak, the number recruited via NRMP in 2007 was larger (by approximately 260) than the number match-recruited as far back as 1992. In 2007, anesthesiology attracted nearly 7% of all US senior medical student candidates in the match, then a historical peak.14 Although interest from US medical students was at an all-time high, the total number recruited into the specialty had leveled off near the Medicare cap for residency positions.15
In academic institutions, upward pressure on anesthesiologists' salaries moderated during this period. The 50th percentile salary paid to an assistant professor increased by only 3.6% in 2005, down from 7.4% in 2004, 8.1% in 2003, and 14.2% in 2002.16 Across the board, salary increases moderated after strong increases in 2000 and 2002 (Table 2a). Except for cardiac surgery, little slowing of the growth in surgical procedures and anesthetizing locations was evident (Table 1). Although there may not have been shortages everywhere, there seemed to be a large number of “regional shortages.”17 During this time, academic anesthesiology departments improved their shortage of faculty from a deficit of 7% to 10% during 2001–2003 to 5% to 7% in 2004–2007 (Table 3).
In 2006, collaboration between ASA and the Association of American Medical Colleges (AAMC) resulted in a 10,000-member survey of anesthesiologists older than age 50 years, with approximately 4000 respondents. Twenty-six percent were women. Thirteen percent of anesthesiologists older than 50 years worked part-time (but only 6% worked ≤20 hours a week) and 67% had no plans to retire at a specific age, whereas 17% planned to retire before age 65 years. Nearly 80% indicated that they would “provide the same or a reduced volume of care in the next three years” (i.e., the 3 years following 2006).18 These observations indicate that a large majority of anesthesiologists will not be increasing their workforce contribution, and possibly decrease it. Taken together with data on desire to go part-time and other trends, it seems likely that workforce contribution of the average anesthesiologist will not increase and will likely decline, thus slowing the growth of anesthesiologist labor supply.
The Ohio Society of Anesthesiologists has conducted membership surveys biennially since 1999. Table 4 shows that the fraction of respondents reporting “position vacancies in their groups” declined by an absolute 10% (or relative 21%), and the fraction reporting a full anesthesiologist roster increased similarly from 2003 to 2007. Likewise, the number of job offers received by New York State anesthesiologists decreased to 3.8 in 2005, after it had been increasing to 4.8 during the years 2000 to 2003.10 In 2004, the electronic jobs bulletin board, www.gaswork.com, listed 57 advertisements for Ohio anesthesiologists, decreasing to 53 in 2006. Similarly, open positions in academic departments had decreased to a mean of 3.3 in 2004 and 2.8 in 2005,16 also reflecting a moderating shortage. This trend was confirmed by the findings of the SAAC e-mail surveys (Table 3) and was reflected in qualitative observations such as the difficulty search firms encounter in recruiting.19 Still, despite the declining average number of positions, the fraction of departments with open positions (87%; Table 3) was higher in 2005 than in each year since 2000, indicating that many academic departments continued to experience difficulty recruiting faculty. Anesthesiologist staffing and recruiting difficulties were also mentioned in 7 of the 50 reports of state society of anesthesiologist reports to the ASA House of Delegates in 2006,20 bearing evidence that regional shortages continued to persist. Further supporting this impression was a 2007 national survey conducted by the RAND Corporation that indicated a persisting, regionally accentuated shortage of anesthesiologists. More than 4000 anesthesiologists and CRNAs each were surveyed, with a 20% response rate, along with 680 anesthesiology directors (>50% response rate). More than 50% of those surveyed reported at least 1 open position; shortages of anesthesiologists were present in 25 of 49 states surveyed, with an estimated shortage of 3800. Shortages were greater in the East than West, and greater with larger employers in urban areas.21 One potential flaw of the RAND methodology is the use of NRMP data averaged over the last decade for calculation of anesthesiologist workforce entry rates. Because NRMP fill rates have increased substantially since 1999–2003, and because NRMP recruitment only represents a fraction of total anesthesiology residency recruiting, the workforce entry rate of 2.5% assumed in the RAND study scenario that results in a shortage by 2020 may have been underestimated, possibly leading to an overestimation of the shortage. We calculated current workforce entry at approximately 4.5% of the active workforce, assuming approximately 1550 trainees entering a workforce of 32,000. However, RAND also assumed demand for anesthesia services to increase linearly during the next decade, rather than exponentially, because it might be expected to parallel growth in the population over the age of 65 years.
There was still an anesthesiologist workforce shortage in 2007, based on the RAND survey, position vacancy data, reports from professional state component societies, as well as salary information. It was regionally accentuated,21 especially in federally defined rural areas where only 7% of active anesthesiologists practiced.22 The percentage of RAND survey respondents reporting “any open positions in your group/practice” for anesthesiologists and CRNAs was highest for urban areas, the Southeast and the Northeast (78%–78% and 60%–72%, respectively, compared with 47%–65% and 40%–54%, respectively, in the rural areas, West, and the Midwest), where they reported that their employers “need more personnel to do more cases.” Primarily because of salary data (Table 2a), we believe that the shortage of anesthesiologists was likely less acute in 2007 than it was in 2001 to 2003, and estimate that it amounted to somewhat less than the RAND estimate, representing a deficit of 2000 to 2500 anesthesiologists.
The fraction of academic departments with open CRNA positions was the highest it had been during any SAAC survey and was 89% in 2005 (Table 5); the average number of CRNA position vacancies had also been increasing for several years and continued unchanged compared with 5 to 6 years earlier. In 2004, the jobs bulletin board Gaswork.com showed 75 position advertisements for CRNAs in Ohio, growing to 121 position vacancy notifications in 2006, suggesting that CRNAs continued to be in short supply. The RAND survey estimated a CRNA shortage of 1186 in 2007. The same year, 54% of US Veterans Administration chief anesthesiologists reported that “some operating rooms were closed temporarily” and 72% reported that “some elective surgeries were delayed” because of CRNA vacancies at their facilities.23 Evidence of a continuing CRNA shortage was corroborated by a 2007 survey of hospital and ASC administrators showing an increase in CRNA vacancy rates from approximately 3300 in 2002 to 5800 in 2007.24 Although survey methods were not strictly comparable, the authors still showed a large increase in vacancies even after correction for survey methodology differences. CRNA vacancies grew faster in ASCs (from 455 to 1196; 263%) compared with hospitals where vacancies grew from 2822 to 4596 or by 162%. There was a marked regional imbalance in CRNA vacancies, with very low vacancy rates in the Great Lakes Region and the Northeast, and much higher vacancies in the Gulf, Southeast, and West. Rural vacancy rates were generally twice as high compared with vacancies in urban areas, except in the Northeast and Mid-Atlantic regions. Based on an AANA membership survey, 15% of CRNAs worked part-time in 2005, and 36% of the then membership of 29,486 CRNAs were expected to retire by 2014,25 suggesting an annual workforce exit rate of 5%, which is substantially above that assumed by RAND. There is a marked discrepancy between the RAND and the AANA workforce shortage estimates for 2007 (1200 vs 5800, respectively). In part, the AANA survey could have overestimated vacancies because of differences in methodology, and because CRNA survey responses may have represented CRNAs' desire to see more colleagues hired, rather than the actual need. At the same time, CRNA workforce exit rates (assumed at 1.4% in the RAND study) likely were underestimated and may have resulted in an overestimation of personnel availability. Given the multiple methods used in the RAND survey to arrive at a conclusion and the “arm's length” position of the RAND corporation vis-à-vis the CRNA versus anesthesiologist debate, we chose to estimate CRNA 2007 shortfall at a point between the RAND and AANA data in Table 4. Although the RAND and CRNA survey results differ in their findings with respect to open CRNA positions in rural areas and in the Northeast, both showed considerable evidence for a continuing need for CRNAs in certain areas, as indicated by reported open positions.
Recent Years of Economic Turmoil: 2008 to 2010
Entry into the specialty of anesthesiology from US medical schools increased further, surpassing a rate of 7% for the first time.14 Recruitment from other training sources, including international medical graduates (IMGs), declined to its lowest level since such records were kept. The number of US anesthesiology residency training programs have not grown, but declined slightly. If this trend continues, the future US anesthesiology workforce will contain far fewer IMGs than is currently the case.
Open faculty positions in academic departments had decreased in 2004 to 2006, but began to increase again, with a continuing 7% shortage of academic anesthesiologists in 2008 and 2009, up from 5% to 6% before (Table 3). As is evident from Table 5, there seems to be a continuing need for CRNAs in academic groups, likely related to ACGME work-hour rules and RRC teaching requirements.
The economic downturn of 2008 has definitely begun to manifest in the health care industry. Starting in late 2008, hospitals began to report reductions in visits and procedures.26 Thirty-eight percent reported a moderate to significant decline in patient admissions, and 31% reported a similar decline in elective operations. Forty-eight percent reported an increase in Medicaid patients, and 58% reported an increase in outstanding receivables. It is likely that patients began to seek access to health care less frequently because of employment uncertainties and/or inability to handle substantive copayments. In addition, federal and state budgets are stressed because of record deficits, resulting in payment reductions. As a result, compensation increases for anesthesiologists have moderated or ceased entirely (Table 2a). In 2010, the search firm Merritt Hawkins saw a decrease in physician recruitment assignments versus the prior year of 14%, the first time in its history.27 Capital budget constraints of health care institutions and the prolongation of retirement by existing practitioners were cited as reasons. During the last 2 years, hiring of anesthesiologists reportedly has been inhibited by reimbursement changes and a decrease in elective and nonelective surgical procedures, prompting us to reduce the assumed growth rate in our workforce model to zero during the years 2009 to 2011 (e-Table 3, see Supplemental Digital Content 3, http://links.lww.com/AA/A395) from 2.5% earlier in the decade.
Demand for Anesthesia Services During the Last Decade
Although we define demand for anesthesia services as a composite of procedures and anesthetizing locations, this likely oversimplifies a concept that is inherently difficult to assess; there is no single quantitative measure of demand for anesthesia work in surgical, interventional, diagnostic, pain management, and critical care settings. The most tangible data are available for surgical procedure and OR growth; we have also summarized qualitative considerations concerning other important factors, citing supporting evidence where available. In our workforce model, we use a composite estimate for anesthesia services, primarily based on surgical procedures and anesthetizing locations, supported by compensation trends, and modified by other factors such as pain and critical care services.
The decade of the 1990s saw an unprecedented expansion of ambulatory anesthetizing locations; hospital-based ASCs grew at a rate of 4% to 6%, while freestanding facilities added ambulatory surgery capacity at a rate of 8% to 9% annually,5 which contributed substantially to the need to staff these new anesthetizing locations with anesthesia personnel. Despite this expansion in the ambulatory surgical capacity, hospitals were also engaging in a building “boom” to remain competitive and attract top talent.28 Anesthetizing locations contribute to the demand for anesthesia personnel even if the increase in cases done at newly created locations does not match the number of new locations. This is because OR management decisions are frequently not made based on maximizing efficiency (or minimizing unutilized but staffed time).29–31
Global Surgical Growth
According to a 2001 analysis by SMG Marketing,5 overall growth rates for surgery were 4% to 5% until the mid-1990s, fueled primarily by a high ambulatory surgical growth rate. In the latter half of the decade of the 1990s, growth rates slowed to approximately 3% (Fig. 1). Surgical growth during the last 10 years has been lower, approximately 1% to 2% annually, with evidence of further slowdowns during the last 12 to 18 months.
According to the Health Policy Institute of the American College of Surgeons,32 the active surgical workforce increased 53% from 1981 to 2006, amounting to approximately 2% per year. The number of surgeons grew at a rate of slightly more than 1% between 2001 and 2006; it was 0.8% during the years 2006 to 2008. This trend confirms a slowing of surgical growth in aggregate, although there is the possibility that each surgeon is now performing more procedures than in the past.
Anesthesia Outside the OR
Interventional, nonoperative procedures, frequently replacing or complementing surgical therapies, have been increasing. It has been estimated that 40% of the anesthesiologist's work is now occurring outside the OR.33 For example, increasingly, pediatric anesthesia has been needed in the cardiac catheterization laboratory.34 A 2005 survey of 116 children's hospitals showed that the most common barrier to the development of pediatric sedation services was a shortage of providers, particularly pediatric anesthesiologists.35 Outside the OR, anesthetizing locations require a higher personnel-to-patient ratio to ensure patient safety in a relatively unfamiliar environment compounded by lower efficiency (when compared with most ORs) and scheduling difficulties. An increasing need for personnel intensive anesthesia services outside the OR suite therefore contributed strongly to hiring of anesthesia providers. Therefore, in estimating demand for anesthesia services, merely considering surgical in-suite case volume is insufficient.
Population Disease Burden
Obesity is associated with joint disease, diabetes, cardiovascular disease, and chronic pain, all conditions for which management by surgeons and anesthesiologists is needed. The US Centers for Disease Control and Prevention has declared the existence of an obesity epidemic, with a prevalence of >30% among the adult population. Adolescent obesity has rapidly become a public health issue; its prevalence has increased from 5% in the 1980s to 11% in 1994 and further to >15% in 2000.36 Obesity contributes to procedural growth by fueling surgical approaches to cure diabetes such as gastric restriction, as well as surgery to correct joint, vascular, renal, and cardiac disease worsened by excess weight and diabetes.37 Approximately 59% of people older than 65 years are affected by arthritis, which is the leading cause of disability in this age group.38 Likewise, 20% of people older than 65 years are affected by diabetes. As the population ages, the total disease burden of conditions such as arthritis, diabetes, and obesity will increase dramatically. For example, by 2050, the prevalence of diabetes will quadruple in people older than 75 years.39 Arthritis drives procedures for joint replacement, spine decompression and stabilization, and interventional pain therapy, all contributing to the need for services provided by nurse anesthetists and anesthesiologists.
Patients older than 65 years incur more than half of the nation's intensive care unit (ICU) days,40 making critical care practice especially sensitive to the aging of the “baby boomer” population segment. In 1997, intensivists provided care to 37% of ICU patients.40 From the results of a large survey of critical care directors and intensivists, Angus et al.40 built a workforce model and reported several sensitivity analyses. For the past decade, most scenarios predicted a supply-to-demand ratio for intensivists close to unity, but decreasing sharply during the next decade (see below). Data from the National Inpatient Discharge Survey41 confirm that patients older than 65 years have an operative procedure rate 267% that of patients aged 45 to 64 years, and 371% higher than patients aged 15 to 44 years. This relationship was even more striking with diagnostic procedures. As patients live longer, they will require health care for increasingly longer periods and require more procedures, most of which necessitate anesthetic care.
State of the Anesthesia Workforce in 2011
Anesthesiologists provide or participate in >90% of the estimated 40 million anesthetics that are administered annually in the US.42 Anesthesiologists represented 4.4% of the total physician workforce in 2008, according to the American Medical Association (AMA) Physician Master File. There were 42,230 anesthesiologists in 2008.43 Of these, 1413 were pain medicine specialists, 431 were critical care specialists, and 4786 were in training. In 2006, according to a survey conducted by the Anesthesia Administration Assembly and MGMA, 21% were in academic and 79% in private practice.44 Based on work by the Center for Workforce Studies of AAMC, 37,134 (or 90%) of all 41,193 anesthesiologists were active in 2006.45 “Active” physicians are defined by AAMC as allopathic and osteopathic federal and nonfederal physicians practicing in the US, who work >20 hours a week and are not in graduate medical training. Excluding trainees, and assuming that the same ratios applied in 2008, there were 33,263 “active” anesthesiologists in 2008, or 89% of the nontrainee workforce. This would imply that 11% of the anesthesiology workforce worked part-time at ≤20 hours per week.
In 2006, 22% or 8759 of the ASA membership were women. This was less than the average of all physicians (27%). In the 1980s, only 15% of anesthesiologists were women. The specialty of anesthesiology has been recruiting proportionately fewer women into the specialty compared with all physician specialties. The proportion of female anesthesiology ACGME residents was 29% in August 2004 compared with an overall average of 42%.46 More recently, female ASA membership, including trainees, has increased to 23.3% or 10,084 women anesthesiologists.42 In 2010, only 22% of actively practicing ASA members are women (Table 2b), suggesting an imminent increase in female practitioners. This trend will hold also for the specialty of anesthesiology given the rapidly increasing fraction of female members of the entering anesthesiology PGY-1 intern classes, from 24% in 2000 to >40% currently.42
The proportion of active anesthesiologists aged 55 years and older was 28%, compared with the average of 33% for all physicians. Only 4 specialties were “younger” than anesthesiology; they are general internal medicine, infectious diseases, physical medicine and rehabilitation, and geriatric medicine. Twenty-nine percent of active anesthesiologists had foreign medical degrees. Of anesthesiologists trained in the US, 46% were practicing in the state where they trained for their specialty.
Nonallopathic anesthesiologists comprise a small part of the anesthesiology workforce. In 2006, 5% of anesthesiologists were Doctors of Osteopathy. Doctors of Osteopathy in ACGME-accredited anesthesiology training programs increased during the early part of the past decade, but have recently leveled off (Fig. 3). In addition to ACGME, the American Osteopathic Association (AOA), through its agency, the Bureau of Osteopathic Specialists, accredits osteopathic anesthesiology residency training programs. Similar to graduates of ACGME-accredited programs, graduates of osteopathic anesthesiology programs achieve board certification and enter the anesthesia workforce, adding to workforce entry by trainees certified by the American Board of Anesthesiology (ABA). Twenty-four AOA trainees are projected to complete residency this academic year (2009–2010). During the last decade, AOA training programs have nearly doubled in size. In 2010, the total number of approved osteopathic anesthesiology positions had increased to 112 (data from the American Osteopathic College of Anesthesiologists). Still, the output from osteopathic-only training programs remains at only <1% of the total annual training output of anesthesiologists. It is conceivable that the higher salaries currently available to anesthesiologists entering the market place might lead to further increasing output from osteopathic-only anesthesiology programs.
Aside from ABA and AOA, there are a few other specialty boards that certify for anesthesia practice. They include the American Board of Physician Specialties, the American Dental Board of Anesthesiology through the American Society of Dentist Anesthesiologists, and the National Dental Board of Anesthesiology through the American Dental Society of Anesthesiology. The American Board of Physician Specialties provides board certification in anesthesiology for allopathic and osteopathic-trained physicians through an application and examination process.43 In 2010, there were 43 active diplomats board certified by the American Board of Physician Specialties (data from the American Board of Physician Specialties). Dental anesthesiology residencies are generally 2-year programs that commence after graduation from dental school. In 2006, there were 7 active programs across the US. Ten to 15 residents graduated from these programs before 2006. In 2010, there were 10 active dental anesthesiology residency programs with 58 resident positions per year. From these programs, 19 dental anesthesiologists graduated in 2010 with a projected graduation of 26 for 2011. These graduates sit for the American Dental Board of Anesthesiology and add minimally to the supply of anesthesiologists, with only 127 currently active diplomats of the American Dental Board of Anesthesiology in practice (data from the American Dental Board of Anesthesiology).
The National Dental Board of Anesthesiology through the American Dental Society of Anesthesiology board certifies dentists and oral maxillofacial (OMF) surgeons for the practice of anesthesiology. Training for the board ranges from a 3-month to 1-year period in anesthesiology during OMF residency or completion of a dental anesthesiology residency. There are approximately 1100 OMF residents training in 103 OMF residency programs in the US approved by the Commission on Dental Accreditation (data from the American Association of Oral Maxillofacial Surgeons). In the US, approximately 85% of the 9000 OMF surgeons practice anesthesia in the office setting. Approximately 2000 are board certified by the National Dental Board of Anesthesiology.
The growth rate in the active anesthesiology workforce from 1997 to 2005 was among the lowest of all specialties, amounting to 17%, or 2.1% annually. Lower growth rates were seen in many surgical specialties (otolaryngology, obstetrics and gynecology, ophthalmology, neurological surgery, urology, orthopedic and general surgery) as well as in psychiatry and pathology.47
According to the AANA, CRNAs “provide anesthesia care for approximately 60% of patients.” There are approximately 30,000 practicing nurse anesthetists and >100 CRNA training programs. Medicare payment rules specify that CRNAs must be supervised or medically directed by a physician. The supervising physician can be a surgeon or anesthesiologist. CRNAs can practice independently, i.e., without physician supervision in the 14 so-called “opt out” states where governors have chosen to opt out of the Medicare supervision rule. Although disputed, the rationale for CRNA independent practice has been that too few anesthesiologists are available for rural practice. Despite several governors recently opting out, a review of state society reports to the ASA House of Delegates fails to reveal practice pattern changes.48
According to the National Commission for Certification of Anesthesiologist Assistants, 1023 anesthesiologist assistants (AAs) are currently certified to practice under delegatory authority of an anesthesiologist in 18 states and in federal hospitals.20 There are currently 7 AA training programs, including at Emory University in Atlanta, GA; Case Western Reserve University in Cleveland, OH; University of Texas, Houston, TX; South University in Savannah, GA; University of Missouri in Kansas City, MO; Nova Southeastern University in Fort Lauderdale-Davie and Tampa, FL, now graduating approximately 140 AAs annually.49 Data from the National Commission for Certification of Anesthesiologist Assistants show approximately 120 new certifications for AAs in 2009 and 2010, substantially increased from 40 to 50 in the years 2001 to 2005. An anesthesiologist may supervise up to 4 AAs or CRNAs. A few states limit supervision of AAs to a lesser supervision ratio.
Critical care fellowship training in anesthesiology increased 23% in the last 2 years compared with the 2 years before. It has held steady during the last 2 years with 62 to 66 fellows currently in training. Of note, the number of anesthesiology critical care fellows graduating is still approximately the same as it was in 1996, when 63 graduated.40 Among anesthesiology fellowships, critical care has the highest vacancy rate. The number of pediatric fellows in 2009 was 151, down from 160 in 2008. Compared with 2006 and 2007, the number of anesthesiologists training in pediatric anesthesia is increased by 24%, however. Graduation from ACGME-accredited adult cardiac anesthesiology fellowship programs increased by 18%, with 99 accredited fellows training in 2009 compared with 84 in 2008. Likewise, in 2009, there were 222 anesthesiologists training in pain management compared with 210 the year before.48,50
Surveys of anesthesiology residency program directors in 2007 and 2008 reveal a differentiation in hiring practices by type of anesthesiology training.11 Sixty-nine percent of open faculty positions in training institutions were for subspecialists. Of those, the highest preference was for pediatric anesthesiologists (30%), followed by cardiac (17%) and critical care (16%). Meaningful anesthesiology subspecialty workforce data are difficult obtain and interpret because of coexistence and conversion of fellowship training programs from nonaccredited to ACGME-accredited status (e.g., cardiac anesthesia), and multispecialty enrollment (pain, critical care). Based on total number in training, vacancy rate, and service volume growth, critical care appears to show the most poignant imbalance between the need for specialty-trained practitioners and interest in subspecialty training.
FUTURE DEVELOPMENTS IN THE ANESTHESIA WORKFORCE
As it has in the past, the anesthesia workforce will be shaped by the supply of labor and growth for service. In the near and intermediate future, many factors will affect labor supply, as defined by work (case work and availability), contributed by actively practicing anesthesia providers. They include graduation rates from residency programs, migration, workforce contribution, and retirement. Workforce contribution (the relationship of clinical work output to total number of active anesthesia providers) is affected by part-time work, partial retirement, migration to nonclinical positions, and migration to other countries. Growth in procedures and perioperative care coordination requiring the services of anesthesiologists and anesthetists is primarily a function of the age and disease burden of our population, modified by resources such as surgical capabilities, payment, and innovation.
Future Anesthesia Workforce Supply
Given the current popularity of the specialty of anesthesiology among US medical students (Fig. 7), a supply of approximately 1500 to 1600 annually (or 4.3%–4.6% of the workforce) is virtually certain for the next 3 years.51 Developments beyond a 3- to 4-year residency training horizon are not feasible to predict. There is the possibility that US medical students may conclude from a temporary pause in growth for anesthesia services that the future of the specialty will be less bright than currently assumed. This may then in turn prompt a decrease in medical students enrolling in anesthesiology residency programs. The RAND study assumes a workforce entry rate of 1.8%, which is substantially below the levels seen currently. This and the impact of the economic downturn have given rise to concerns that the RAND analysis might overestimate the projected anesthesiologist shortage in 2020. Two factors that mitigate such concerns are the likely underestimation of retirement rates (see below) and the possibility that residency program enrollment might decrease, as indicated above.
Today's generation of medical students base their career choices more on perceived lifestyle and income than in the past. Controllable lifestyle explained 55% of the variability in specialty preference from 1996 to 2002 after controlling for income, work hours, and training length. Perception of controllable lifestyle accounts for most of the variability in recent changing patterns in the specialty choices of graduating US medical students.52 In this study, anesthesiology was classified as a specialty with controllable lifestyle, despite being highest in work hours.50 After lifestyle, income was second in accounting for medical students' career choices. As the newer generations become the predominant element of the workforce, women physicians will no longer delay childbearing, resulting in less workforce contribution from maternity/paternity leave and part-time work patterns. Emphasis on a healthy work-life balance will likely prevent increases in hours worked per week, limiting workforce contribution and labor supply. Residency duty hour restrictions have become even more stringent in 2011 (www.acgme.org). With a relatively fixed training output and limited ability of the new generations to increase work contribution, the anesthesiologist workforce of the future likely will not perform the same work complement as in the past. The current workforce supply system therefore cannot be relied on to meet the increasing procedural and cognitive work required of anesthesia providers in the future.
With US medical schools graduating >50% female physicians, anesthesiology residency program entry is becoming increasingly female as well. The fraction of women entering the PGY-1 class was 40% in 2009. It has increased to 39% in the CA-1 class, up from 38% in 2007 and from 28% in 2000. If sustained, this trend will increase the proportion of female anesthesiologists in the workforce, whose work contribution is approximately 10% less than their male counterparts. Year for year, this effect is relatively small, given the slow pace at which workforce composition changes. At the current rate of production and retirement, the workforce should gain 1.2% to 1.4% women each year (calculation based on 36,000 active anesthesiologists, 1600 workforce entrants, and 800 workforce exits). Given a 10% lower work contribution compared with the men they will be replacing, maintenance of total work supply would require a compensating increase of 0.12% to 0.14% in labor supply.
Retirement and Mortality
Retirement is an important factor in the workforce,51 yet data are difficult to obtain. Based on the ASA-commissioned Tarrance Group survey of 2363 US hospitals,8 and assuming that only 1 anesthesiology group served each large hospital, one arrives at an estimate of approximately 800 anesthesiologist retirements (2.3% of an active workforce of approximately 34,000; Table 2b) per year.4 Given the age distribution of ASA members (Fig. 4) and a retirement age of 65 years, ASA member retirements in 2002 should have been approximately 400. Because not all anesthesiologists retire at age 65 years, nor do ASA members, our workforce model estimates total anesthesiologist retirements from 80% of proportion of ASA members and active anesthesiologists (from the AMA Physician Master File), amounting to an annual retirement rate of approximately 2%. This workforce exit rate is augmented by deaths, which have been estimated at 1%,21 making for a workforce exit rate of 3%. Based on ASA active membership data, annual retirement rates should begin to increase to as high as 1000 anesthesiologists annually by 2017, as the cohort now 47 to 55 years old reach retirement age (Fig. 4). As the current workforce ages, in particular the large group of baby boomers, workforce contribution is reduced as anesthesiologists reduce their work hours near retirement. However, this phenomenon may be counteracted by retirees reentering the workforce on a part-time basis, especially when economic conditions deteriorate. The RAND study uses a workforce exit rate of 1.82%, suggesting a very conservative estimate. If, as we suggest above, the exit rate assumed by RAND is lower than the actual exit rate, the shortage of anesthesiologists in 2010 may be more severe than enumerated in the RAND report.
Labor substitution is a potential and in some areas a real factor influencing the anesthesia labor market. Anesthesia can legally be administered by a number of different provider groups. Substituting for anesthesiologists may be CRNAs, AAs, family practitioners, oral surgeons, etc. In 2001, the Centers for Medicare and Medicaid Services issued a ruling that allowed state governors to opt out of the medical supervision rule for Medicare reimbursement. It seems that CRNAs have been able to expand their scope of practice somewhat in both opt-out and non–opt-out states. Based on Medicare parts A and B claims from the years 1999 to 2005, CRNA-only (or “solo”) care increased from 18% in 1999 to 22% in 2005, whereas anesthesiologist-only and anesthesiologist-CRNA team care decreased from 82% to 78% in opt-out states. In non–opt-out states, solo CRNA claims increased from 7% to 12%, and anesthesiologist solo or care team claims decreased from 93.1% to 88%.53 Still, such data are not conclusive as claims suggesting CRNA solo practice may be submitted by anesthesiology groups for the sole reason of avoiding onerous regulatory requirements for medical direction. Moreover, physician-supervised anesthesia is thought to be cost-effective compared with a CRNA intensive care model, conceivably limiting an economic driver of substitution.54
Although AAs could eventually practice independently, their current charter requires supervision by an anesthesiologist. Furthermore, their numbers are few, accounting for only a few percent of the nonphysician anesthesia workforce. Besides CRNAs, a number of other providers practice anesthesia. Aside from the ABA and the American Osteopathic Board of Anesthesiology, there are a few other specialty boards that board certify for anesthesia. Those include the American Dental Board of Anesthesiology through the American Dental Society of Anesthesiology, the American Board of Physician Specialties–Diplomat in Anesthesiology, and the National Board of Anesthesiology. These groups are too small to provide a significant labor alternative for providing anesthetic care.
Relationship Between Compensation and Demand for Anesthesia Services
There is evidence of a direct relationship between compensation and the number of anesthetizing locations (Tables 1 and 2a), which we have previously identified as a major influencer of labor demand. First, RAND survey data analysis reveals higher anesthesiologist compensation rates in states with shortages, with a premium of 10% to 12%.21 Second, AMA's physician socioeconomic data show that a wage (or compensation) decrement of $10/hour (approximately $20,000 annually) increases the proportion of retired anesthesiologists by 0.7%. As compensation decreased, fewer anesthesiologists opted to continue to work and chose instead to exit the active workforce, illustrating a direct relationship between compensation and supply of labor.51 Third, hospitals have had to support increasing anesthesiologists' compensation in response to growth in anesthetizing locations and the inability of labor supply systems to fulfill this demand. Fourth, at a time when faculty vacancy rates continued to be 7% to 10% (Table 3), per-faculty wage supplements to academic anesthesiology departments increased by 75% >2 years, and overall anesthesiologist compensation increased 7.9% (Table 2a). In contrast, when faculty vacancy rates decreased to the 5% to 6% range in 2004 to 2006 (Table 3), annual hospital supplementary compensation increases remained <12% (Table 2a). Between 2005 and 2006, hospital supplementation of anesthesiologists' compensation increased only 3% from $116,000 to $120,000 per faculty member; the figure now stands at $165,000 (Table 2a). Fifth, an annual increase of 2% to 3% in surgical procedures and ORs during the past decade was associated with compensation increases of 5%, suggesting a continued shortfall in anesthesiologists. Demand for anesthesia service not captured by facility OR data (Table 1), such as that associated with pain treatment and out-of-OR procedures, might also have grown more rapidly, making the true growth for anesthesia services exceed 2% to 3%.
These relationships support our contention that compensation trends indicate continued demand for anesthesia personnel. The continuing increase in compensation rates indicate, at a minimum, that there have been and continue to be relative shortages of anesthesia providers given the increasing number of anesthetizing locations (Table 2b). Because we consider trends in compensation an indicator of demand for personnel, we maintain that there must be a shortage to generate this type of compensation response by hospitals and employers. To explain further, a shortfall in providers leads to greater difficulty in recruiting, which again results in groups or facilities offering higher salaries to attract new anesthesiologists or incentivize those on staff to increase their workload.
Although one might argue that the need for anesthesia services could not have increased if hours of clinical care per provider per day stayed constant, practically speaking, the latter measure does not capture market dynamics in many anesthesia labor markets. As is evident from Table 2b, we consider that the numbers of anesthetizing locations are a vital and important driver of demand. Hospitals and other operating facilities are competing for surgical business in part on availability of staffed OR time, often initially without regard for OR utilization, defined as the ratio of actual case time to staffed time in any anesthetizing location. Because surgeons are influential in directing case volume to the facility, expect early start times, and rarely will voluntarily consent to follow another surgeon, the need to staff these new locations has resulted in groups having to hire additional providers. This has occurred even in the face of stable traditional measures, such as ASA units or cases per anesthesiologist per year (Table 1). Data from the RAND study tend to corroborate this notion, because labor rate premiums were observed in states with anesthesia personnel shortages that were based on economic, but not epidemiologic, demand-based analyses.21
Future Demand for Anesthesia Services
Demand for anesthesia services is defined by several factors, chiefly the number of procedures requiring anesthesia, the number of anesthetizing locations, as well as the extent of pain and critical care services. Demand for service of anesthesiologists and other providers will still be influenced by surgical growth in the future, but increasingly also by interventional and diagnostic needs. The 2010 health care reform legislation and other substantive health care policy changes will likely have a profound effect on the number of procedures as more Americans obtain health insurance.
Future Global Surgical and Procedural Growth
Comprehensive data quantifying the expected growth of all surgical procedures in the US are not available. Global inpatient hospital volume is forecast to increase at a rate of 1% annually, whereas outpatient volume is expected to increase by 1.8% annually through 2016.55 According to a 2000 marketing analysis,5 global growth rates for surgery were predicted to remain at approximately 2% after the year 2000. Based on National Discharge Hospital Survey from the National Center for Health Statistics (NCHS), hospital inpatient procedure growth amounted to approximately 1% per annum during the years 1996 to 2005.56 NCHS estimated in 1996 that, in addition to the >40,000 inpatient procedures, another 31,600 outpatient procedures were performed in ASCs. Approximately two-thirds of these were “operations”; it is unclear how many required anesthesia personnel. The 2005 NCHS inpatient data reconfirm that patients 65 years and older require >3 times the number of procedures compared with the general population. Therefore, as the US population ages, procedure growth may well continue to accelerate. From 2010 to 2030, the number of Americans aged 65 years and older will increase by 28 million from 40 to 68 million. By comparison, this same cohort increased by only 6 million during the previous 20 years (US Census Bureau, Decennial Census Data Population Projections, 2010). In 2003, Etzioni et al.41 published an analysis of future surgical workload using recent US Census data, as well as data from the 1996 National Hospital Discharge Survey and National Survey of Ambulatory Surgery. The authors included 7 specialties and 214 representative surgical procedures. They calculated future workload from present age-specific incidence rates for each procedure multiplied by the physician work component relative value unit. Each surgical specialty had 13% to 15% of the total workload. Workload increases were in the range of 30% to 35%, which corresponds to annual workload increases of 1.3% to 1.5%. The estimate of future surgical procedures by Etzioni et al.41 was based on data sources that do not include freestanding surgical practice, nonsurgical anesthesia work, pain management work, or critical care work. An estimate of 1.5% annual growth for anesthesia services still seems conservative, considering a higher recent service growth assumption.21 These estimates do not take into consideration that a surgical workforce shortage has been forecast with a need to train an additional 100,000 surgeons between now and 2030.57 If surgeons cannot be added, surgical growth rate may be constrained or care models may change. Population aging and health care usage trends were considered in a 2010 workforce analysis for anesthesiology conducted by RAND Corporation and yielded an annual service growth rate of 1.6%.21 After using anesthesia service growth estimates of 2% to 3% in the 1990s through 2002, our model has adopted a 1.5% growth rate until 2008 where we inserted a “growth pause” based on economic reality. Our model projections assume anesthesia service demand growth in 2012 increasing from 1% to 2.5% in 2020.
Presently Unmet Patient Needs
There are still many unmet needs in US health care that require anesthesia services. More than 30 million Americans now live in federally designated Health Professional Shortage Areas; at least some will require subspecialty anesthesia services. Epidural anesthesia is available to only 51% and 17% of urban and rural parturients, respectively.58 Clearly, more women would like to take advantage of this service. It is estimated that >75% of ICUs in the US do not have a full-time intensivist and that only 33% of ICU patients have the benefit of a critical care specialist.59 This is essentially unchanged from 1996.40 The need for ICU care is expected to increase with a rapidly aging population, while at the same time, critical care specialists are not increasing.60
Future Subspecialty Needs
Surgical procedure increases will likely differ by subspecialty (Table 6). The highest growth specialties are projected to be ophthalmology, cardiac surgery, urology, and surgical oncology.57,61 Some centers have experienced decreasing or stagnant cardiac surgical case volumes, although innovative minimally invasive procedures are growing. Anesthetic management for these innovative cardiac procedures is often more complex than for standard coronary artery bypass surgery. Projections by Pennington62 and Etzioni et al.61 indicate that the need for cardiac surgery will continue to increase, although the types of surgical interventions will be different, favoring valve surgery, Maze procedures, minimally invasive procedures, and assist and restraint device placement.63 Solid organ transplantation has increased by 21% over the past decade as reported by the Scientific Registry of Transplant Recipients.64 Anesthesiologists specializing in liver transplantation are sought after and cannot be replaced easily. Given the current limited access for the majority of critically ill patients to critical care specialists, the increasing population of aged patients with critical care needs, and the stagnant output of critical care physician training programs, the need for more critical care specialists will continue to increase. Predictions by Angus et al.40 call for a shortage of at least 1000 intensivists by 2020, but more likely closer to 5000, or 22% of total needs. A more recent analysis projected a 38% increase in need for critical care services by 2010 accompanied by an increase in intensivist supply by 48% from 1900 in the year 2000 to 2800 in 2020. This assumes that only one-third of critically ill patients will continue to be cared for by intensivists. If two-thirds of patients could benefit from specialist-provided critical care, the number of intensivists required would increase to 4300, creating a shortfall of 1500 by 2010.59
During the last 18 to 24 months, centers have seen decreasing or stagnant cardiac surgical volumes, although innovative minimally invasive procedures are growing. In addition, many have seen a broader decrease in elective procedures26 as patients delay health care expenses to conserve cash and minimize workplace absence in an economy characterized by high unemployment. It is likely that, over the short term (1–2 years), a temporary plateauing of surgical procedures occurs as the effects of continued high unemployment persist. Despite the economic downturn, in a nationwide survey conducted in 2009 by the medical search firm Merritt Hawkins,27 85% of hospital chief executive officers (CEOs) characterized the search for new physicians to be “somewhat or very difficult,” indicating an average physician vacancy rate of 11% at their hospitals. Again referring to all physicians, 95% of hospital CEOs indicated that there continues to be a moderate or severe physician shortage. This survey did not, however, specifically address anesthesiologists.
Productivity and Innovation
Clinical productivity is a potentially powerful factor affecting workforce supply and secondarily the need for anesthesia providers. There are 2 aspects to clinical productivity of health care personnel. The first is related to workforce contribution or hours worked per full-time equivalent. It is measured by an anesthesiologist's billed hours or available hours per clinical work day. It can increase as anesthesiologists work longer days, or increase their concurrency ratio, i.e., work to cover more anesthetic cases with CRNAs or AAs. Because of the peculiarities of payment for anesthesia services, financial productivity (ASA units billed per clinical work day) increases with decreasing case length and case complexity. It is reasonable to assume that most anesthesia practices have gone through a productivity enhancement stage such as that described by Freund and Posner.63 A 7% increase in clinical productivity from the period of 1994 to 1996 to the period of 1998 to 2000 was accompanied by a longer workday. Given the generational, gender, and age trends in the US anesthesia workforce, it is unlikely that productivity based on workforce contribution will increase. True clinical productivity increases would result if work output, for example, as measured by ASA units or relative value unit, increases per time unit of work. Recently, published data indicate a trend to the contrary, with the number of ASA units per faculty member staying the same or showing a slight decrease during the years 2004 to 2008 (Table 1).11,65 Such productivity is limited by restrictive regulations and poor scheduling practice, and can improve with efficient scheduling, information technology, process, and facility modifications. For example, parallel processing in the OR can result in decreased turnover time,66 resulting in fewer unbilled hours for the anesthesiologist. The future might also bring alternative care models, wherein an anesthesiologist might care for many more patients simultaneously, accompanied by increased safety of medications and surgical interventions. Increasing sophistication in communications technology, or telemedicine, has already extended the reach of intensivists to smaller hospitals where intensivists are not on site. Although all of these developments could affect demand for anesthesia providers, predictions cannot be based on them absent of definitive data.
Health Care Policy
The Healthcare Reform Legislation of 2010, and public sentiment in general, call for and require greater cost efficiency in US health care. The cost of anesthesiologists continues to increase, as seen from Table 2a, as well as the supplement that hospitals pay anesthesia departments per faculty member.11,65,67 If one is to believe that health care policy matters, several scenarios could unfold. In the “Strong Demand” scenario (Fig. 5, Table 7), 32 million additional users of health care will require services phased in over the years 2014 to 2019, including anesthesia services. Recent estimates based on health status and physician access rates for the uninsured indicate that expanded coverage will approximately double physician requirements for the uninsured. c With the supply of anesthesiologists only increasing slowly, a further escalation of salaries would result. This is plausible considering that 70% of hospital CEOs believe they do not have a sufficient number of physicians to handle the additional procedures created by universal access to health care.27 Furthermore, the new health care law is not expected to reduce health care spending as a percentage of gross domestic product.68 As fewer per-capita health care dollars become available under health care reform with more patients accessing service at very low payment rates for anesthesia services (i.e., at Medicaid rates), physicians will perceive they would have to accept lower wages, resulting in decreased labor supply. In response, health care facilities or nongovernmental payers may see an even greater need to cross-subsidize physician salaries or shift to lower compensated providers such as CRNAs, perhaps resulting in substantial compensation growth for these providers. At least for CRNAs, this compensation scenario seems unlikely given the rapid expansion of CRNA training program output and the predictions by RAND Corporation of a surplus of 10,866 CRNAs in 2020 with a 40-hour work week.21 Still another alternative to the Strong Demand scenario is the scenario of a “Disruptive Innovation,” as discussed by Sandberg.69 Examples of such disruptive innovations might be novel noninvasive therapies that do not require anesthesia, or safe, easily administered medications that substitute for an anesthetic and either lessen or completely obviate the need for trained anesthesia providers.
SYNTHESIS AND DISCUSSION
A better understanding of the evolution in the anesthesia workforce and the balance between supply and demand for anesthesia personnel may have several salutary consequences. First, predictive accuracy of projections in labor supply and service growth may improve. Furthermore, such knowledge may be able to better influence the behavior of prospective entrants and employers so that large perturbations in workforce entry are avoided. The current economic downturn presents such a risk. Without proper and comprehensive in-depth analysis of the anesthesia workforce, prospective entrants (mostly medical students) could be persuaded to look for other specialties, given the temporary softening in demand for anesthesiologists. By the time they would have completed training 5 years from now, however, the economy will likely have recovered. By then, the effects of the underlying limited supply characteristics of the US anesthesiology workforce coupled with the anticipated growth in services will likely again result in a situation where there is too much work for the existing cadre of anesthesiologists. Either there will be rationing of services or providers will work more, with ensuing compensation increases. Either way, this situation will have the flavor of a shortage, albeit in the context of the new realities of progressing health care reform.
It has been difficult to project how many residents will graduate beyond a 3- to 4-year time horizon, i.e., the length of the residency training period. Table 8 shows a comparison of the residency graduation cohorts foreseen by our model compared with the actual graduating class sizes from 2001 to 2010. For the years 2002 to 2004, projections were fairly accurate, but they overestimated workforce entry subsequently. This is in large part attributable to the Medicare “caps” on the total number of funded residency positions and the inability for academic departments to increase residency training positions fast enough. Nevertheless, some expansion of residency training slots did occur during those years, albeit slowed by RRC approval procedures and increasing financial burdens on departments and their medical centers. In the years 2003 to 2005, the number of actual graduates exceeded projection. The model subsequently called for 1600 graduates, which was not realized until 2011.
In modeling supply and demand for anesthesiologists during the past decade, we assumed that the prevailing health care policy and economic climates would continue. We extrapolated service growth using 1.5% to 3% yearly growth rates based on a synthesis of recent and projected procedure growth rates, procedure rates for the elderly and population aging trends, and estimated supply based on projected graduation and retirement rates. We assumed that after 2003 the number of American medical graduates would first increase by 15% and that IMGs would decrease to a stable level of 500 trained each year. In fact, recruitment of graduates of US medical schools increased at a much higher rate, with fewer than 200 IMGs being recruited annually now. This development boosted the availability of workforce entrants because fewer J-1 visa holders graduating from US residency programs were lost because of the requirement to return to their home countries. The fact that substantial shortages persisted through 2007 suggests that surgical and procedural needs grew at the higher end of the originally projected 1.5% to 3% range, so that we revised our 2010 workforce model accordingly, assuming anesthesia service growth of 2% to 3% until 2002.
Beginning with the year 2000, entry of anesthesiologists into the workforce increased markedly, continued to increase during the years 2002 to 2006 and began to level off thereafter (Fig. 2). Our predictions that annual training program output needed to increase to approximately 1600 to keep up with demand for anesthesia services by 2010 has been borne out, given the graduation rates of >1550 in 200950 and 2010.70 As mentioned before, our model's assumed growth rate correctly predicted a substantive shortage of anesthesiologists in the early part of the decade. Based on the RAND survey service demand estimate of 1.6%, a 5.3% annual anesthesiologist compensation growth (Table 2a) and the 3.1% growth in the number of ORs (Table 1), we conservatively estimated service growth at 2% early in the decade, and 1.5% from 2003 to 2006.
To account for the recent economic downturn, we adjusted our model to reflect slowing of growth in 2007 and 2008, followed by a near-zero service growth rate for the years 2009 to 2011, followed by a gradual resumption of growth. Figure 5 illustrates the output from our revised workforce model (e-Table 3, see Supplemental Digital Content 3, http://links.lww.com/AA/A395) showing a moderating shortage until 2008. From 2008 onward, in part resulting from a pause in growth, a relative balance in supply and demand results for the present time and continues until 2013. Thereafter, with projected growth in anesthesiology services again increasing (to 2.2%), based on population and policy factors, there is a gradual reaccumulation of an anesthesiologist shortage, the extent of which is captured in 2 likely service demand scenarios (Table 7; scenarios I and II).
What should be done to prevent a substantive anesthesia workforce supply imbalance, that is, shortages and surpluses, from occurring in the future? Our workforce model is primarily based on anesthesiologist supply and demand, and treats other providers conceptually as factors determining demand for anesthesiologists through their capability for extension or substitution of anesthesiologist-provided services. There is no doubt that health care reform places greater emphasis on the use of nonphysicians to grapple with increasing access to medical services and cost. Accordingly, CRNAs, AAs, and other personnel, such as specially trained sedation nurses, will need to be considered in an optimal solution for the most value-based anesthesia care model. A cost-benefit approach may be helpful to further increase clarity over how this could be accomplished.71
To gauge the opportunity to avoid anesthesia labor market imbalances, an assessment of the magnitude of solo anesthesia services seems in order. A published analysis of the Medicare parts A and B claims 1999 to 2005 limited datasets provides some insight.72 The fraction of Medicare surgical anesthetics provided by solo anesthesiologists was 53% in 1999 and deceased to 48% in 2005, whereas use of team-based care models did not change. Theoretically, an opportunity to avoid an anesthesiologist shortage might accrue from fewer personally performed anesthetics. Across all states, this amounted to approximately 620,000 Medicare anesthetics. MGMA73 reports 937 to 965 patients per solo anesthesiologist practicing in 2005. Assuming 950 anesthetics per solo physician and a 17% Medicare market share, >3800 anesthesiologist full-time equivalents were consumed providing solo physician anesthesia care. Theoretically, 2500 of the shortage of 4500 anesthesiologists predicted by RAND and 4000 predicted by our future Strong Demand scenario (Table 7) could then be accommodated if these solo practitioners could convert their practice to a care team model with 1:3 supervision.
The same publication offers limited evidence for labor substitution from CRNAs working independently.52 During this period, the percentage of surgical anesthetics billed as if a CRNA performed the service solo or supervised by a surgeon only increased from 7% to 12% in non–opt-out states and from 18% to 22% in opt-out states. Non–opt-out states accounted for 7 times the number of surgical hospitalizations compared with opt-out states. Adjusted for case volume, therefore, 9.5% of the 1999 surgical anesthetics analyzed were provided under this payment mechanism, increasing to 15% in 2005. Nominally, this represents an annual increase of 0.95%. This should be considered a maximal value as anesthesiology groups have increased their use of this same billing mechanism to mitigate the impact of increasingly restrictive regulation.
For the years 2002 to 2006, the need for anesthesiologists was conservatively calculated as the difference in procedure growth (approximately 3%), and a putative substitution loss of 1%, yielding an anesthesiologist-specific demand growth of approximately 2% (e-Table 3, see Supplemental Digital Content 3, http://links.lww.com/AA/A395). Anecdotally, during the last several years, hiring of anesthesia personnel has slowed, as evidenced by many departments having fewer difficulties recruiting; moreover, a recent survey reported that non–primary care residents had greater difficulty in finding a position and fewer job offers.74 Validating the relationship between demand for anesthesiologists and compensation, both academicd and overalle compensation rates have declined slightly in the last year. Accordingly, our model progressively decreases demand for anesthesiologists leading into the economic downturn beginning with 2008.
The 2007 comprehensive RAND survey of anesthesiologists, CRNAs, and hospital administrators used a combination demographic and econometric analysis to predict a shortfall of 4500 anesthesiologists in 2020, assuming a 49-hour work week.21 Our revised predictive model, using a 2.2% future anesthesiologist-specific service growth assumption (e-Table 3, see Supplemental Digital Content 3, http://links.lww.com/AA/A395), forecasts a somewhat smaller shortage of approximately 4000 anesthesiologists in 2020 under Demand Scenario I (Strong Demand; Table 7). Current growth (excluding the effect of the economic downturn) has been estimated at approximately 1.6% by RAND. Future growth in anesthesiology services is very likely to be higher. We propose that 1.8% is a reasonable and conservative estimate based on population aging and disease burden (Table 6), considering the 1% annual aggregate population growth rate projected by the National Census Bureau, the 2.7% annual increase in the population of patients older than 65 years, the annual 3% increase in case mix index (Table 1), and annual subspecialty growth rates ranging from 1.2% to 2.4% (from Table 6). The effect of health care reform on aggregate need for anesthesia services is not yet foreseeable. However, preliminary forecasts on the effect of demand for physician services have been prepared and indicate that the anticipated US physician shortage will increase from 30,000 to 58,000 in 2014 and from 64,000 to 91,000 in 2020.75 By 2020, the service portion that physicians will need to allocate to accommodate the previously uninsured (9%–10% of the US population) will double from 4% to 8%, suggesting a service demand increment of 0.4% annually over the next 10 years.76 Together with the population, age, and disease-based anesthesia service demand increase of 1.8%, the anticipated service growth associated with expanded coverage under health care reform yields our 2.2% projected growth rate.
Demand Scenario II (“conservative growth”; Table 7) accounts for a continued annual substitution rate of 0.9%, as developed above. This scenario reflects a 41% reduction in anticipated need for anesthesiologists' services compared with a situation without continuing provider substitution under health care reform. With these assumptions, our model still yields a shortage of approximately 1000 anesthesiologists by 2020. This scenario is provided because overall demand for anesthesia services is likely to increase, but continuing growth for anesthesiologist-provided services may be mitigated by ongoing substitution by CRNAs and other providers. However, readers should understand that the data sources for these projections are not specific to anesthesiologists, are not yet peer reviewed, and do not include potential additional demand for service from the currently underinsured who represent as much as 20% of the US population.77 CRNAs have been associated with the care of healthy patients (defined as ASA physical status I and II).78 They also provide a far larger proportion of rural anesthetics than anesthesiologists.79
As case complexity increases (case mix index; Table 1) and the population ages, the number of ASA physical status I and II patients as a proportion of the entire population is likely to decrease, limiting the potential for continuing substitution of anesthesiologists with CRNAs. Similarly, rural counties are generally growing at a lesser pace than suburban or urban areas, further limiting CRNAs' practice growth potential with respect to these practices. Although it may be safer for CRNAs to anesthetize healthy patients, compared with anesthetizing patients with severe or life-threatening illness (ASA physical status III, IV, or V), there is currently some evidence to indicate differences in outcome between anesthesia care models. For example, risk of mortality and failure to rescue were higher in patients whose CRNAs were not directed by an anesthesiologist.80 Furthermore, analysis of physician functions has not confirmed that all physician functions are substitutable; in fact, only maximally 83% were.81 We therefore consider Demand Scenario II a less likely view of the future compared with Demand Scenario I.
To provide a lower bound for estimating workforce balance in 2020, we calculated a very conservative scenario based on a constant 1% growth rate, which is below any reported service growth assumptions. Under such a scenario, accountable care organizations, clinical care integration, and shared savings programs would flourish, destroying the need to perform procedures by >50% compared with Demand Scenario I. In this unlikely scenario, a surplus of nearly 3000 anesthesiologists would be seen by 2020.
Training output for CRNAs has increased unabatedly because of a substantial increase in training positions amounting to approximately 150 additional graduates annually during the past decade (Fig. 6). This is in contrast to the plateauing trend seen for anesthesiology residents' training output (Fig. 2). Increased CRNA training output has continued despite the pause in service growth associated with the economic downturn, and accounts for the RAND-projected CRNA surplus by 2020, an assessment made even before the full impact of the economic depression. If current conditions prevail into the future, a surplus of nearly 8000 CRNAs is predicted by 2020. A continuing 0.9% substitution effect of CRNAs for anesthesiologists would accommodate another 3000 positions for CRNAs, or 300 annually. The figure 3000 of is the difference between Demand Scenarios I and II from e-Table 3 (see Supplemental Digital Content 3, http://links.lww.com/AA/A395) and Table 7. If solo anesthesiologists converted 100% of their practice to 1:3 medical direction, an additional 2500 CRNAs could be accommodated, substantially mitigating, but not eliminating the predicted surplus. From a value-based perspective, one might suggest that anesthesiologists should continue their closer attention to ASA physical status III and IV patients, which are approximately 34% of patients in the National Anesthesia Clinical Outcomes Registry of the Anesthesia Quality Institute. If 34% of solo anesthesiologist–performed cases were medically directed 1:2, the number of additional CRNAs needed would be 2300. Many practical challenges would make even a smaller practice conversion daunting.71 Substantive public policy changes will be needed to sway such major change in anesthesiologist practice patterns. If no further action is taken, CRNA supply is expected to eventually decrease from lower interest of nurses to enter CRNA training as fewer positions become available. Absent policy change, CRNA training programs should limit output until more definitive data are available that account for the effects of the recent economic downturn and the impact of health care reform.
Study of the US anesthesia workforce yields important observations. Anesthesiologist supply has increased steadily, and is at an all-time high, albeit with a plateaued training output. A shortage of anesthesiologists available to contribute actively to the workforce is likely to occur in our 2 most plausible service demand prediction scenarios. Training output of CRNAs has increased rapidly during the past decade. Given the high rate of CRNA production, and the effect of current economic conditions, supply of CRNAs is beginning to exceed the need for CRNAs, a trend that will likely persist in years to come.
A temporary mitigation in demand for anesthesiologists' services is occurring presently. This is attributable to the recent economic downturn. Higher demand for anesthesiologists, who contribute actively to the workforce, is likely to resume with improved economic conditions. Only if the practice conversion from solo anesthesiologist to medical direction practice occurs at a >50% completion level will there be a surplus of anesthesiologists under our Conservative Demand scenario. No amount of solo anesthesiologist practice conversion will be able to eliminate the anesthesiologist shortage under the stronger demand scenario, which we consider more likely. Despite the economy, powerful trends such as the graying of the population, an increasing disease burden, and health care reform will continue to fuel increases in surgical and interventional procedures, critical care, and pain treatment, favoring the employment of anesthesiologists over CRNAs, whose practice is associated with care of healthier patients.
To avoid a CRNA surplus, medical direction models should be encouraged and CRNA production slowed. Shared care integration and savings programs, already part of the health care reform package, may discourage facilities' desire to have open access for surgeons' convenience despite underutilization (e.g., a ratio of case hours per staffed interval per anesthetizing location of <80%). Beyond destroying unlimited surgical access to operating facilities, these mechanisms likely will not translate into more efficient anesthesia care, given current regulatory limitations on anesthesia supervision, coding, and billing practice. Hence, policy makers should consider changes in payment mechanisms that encourage more efficient perioperative care models in appropriate settings, including supervision of specially trained, but less costly personnel.
Furthermore, anesthesiology training programs should be incentivized to train more anesthesiologists who can viably practice in these more efficient environments with an emphasis on practice in geographic and specialty areas where shortage conditions are most severe.
Name: Armin Schubert, MD, MBA.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Armin Schubert approved the final manuscript.
Conflicts of Interest: The author has no conflicts of interest to declare.
Name: Gifford V. Eckhout, MD, MBA.
Contribution: This author helped write the manuscript and Manuscript Review.
Attestation: Gifford V. Eckhout approved the final manuscript.
Conflicts of Interest: The author has no conflicts of interest to declare.
Name: Anh L. Ngo, MD, MBA.
Contribution: This author helped design the study and write the manuscript.
Attestation: Anh L. Ngo approved the final manuscript.
Conflicts of Interest: The author has no conflicts of interest to declare.
Name: Kevin K. Tremper, PhD, MD.
Contribution: This author helped Manuscript Review.
Attestation: Kevin K. Tremper approved the final manuscript.
Conflicts of Interest: Kevin K. Tremper consulted for GE Centricity and received royalties from GE Centricity. Dr. Tremper's department is the development site for GE Centricity, an anesthesia information system.
Name: Mary D. Peterson, MD, MHA.
Contribution: This author helped Manuscript Review and Data Source Identification.
Attestation: Mary D. Peterson approved the final manuscript.
Conflicts of Interest: The author has no conflicts of interest to declare.
This manuscript was handled by: Franklin Dexter, MD, PhD.
We thank Deven Kothari, MD, and Shiva Birdi, MD, for their contributions to the manuscript and Tiffany Hess, BA, as well as Tanya D. Smith, BA, PA (editorial assistant), for their excellent work in resource acquisition and manuscript preparation.
a Anders G. Once Hot Specialty, Anesthesiology Cools as Insurers Scale Back. Washington, DC: The Wall Street Journal, March 17, 1995, p 1.
b These organizations are now part of the Society of Academic Anesthesiology Associations (SAAA).
c See https://www.aamc.org/download/185578/data/2011_pwc_turner.pdf.
d SAAC Survey National Report, personal communication from Rebecca Lovely, University of Florida.
e Data from MGMA Physician Compensation and Production Survey books for years 1990–2010 for the private sector and academic departments. Median income is from Table 1 (private) and Table 3 (academic). Adapted with permission from MGMA.
1. Estimation of Physician Work Force Requirements in Anesthesiology. Bethesda, MD: Abt Associates, Inc., 1994:1–53
2. Schubert A, Eckhout G, Tremper K. An updated view of the national anesthesia personnel shortfall. Anesth Analg 2003;96:207–14
3. Mastropietro CA, Horton BJ, Ouellette SM, Faut-Callahan M. The National Commission on Nurse Anesthesia Education 10 years later. Part 1. The commission years (1989–1994). AANA J 2001;69:379–85
4. Schubert A, Eckhout G, Cooperider T, Kuhel A. Evidence of a current and lasting national anesthesia personnel shortfall: scope and implications. Mayo Clin Proc 2001;76:995–1010
5. 2000 Guide to Healthcare Market Segments. Chicago: SMG Marketing Group, Inc., 2000
6. Fallacaro MD, Ruiz-Law T. Distribution of US anesthesia providers and services. AANA J 2004;72:9–14
9. Cooper RA, Stoflet SJ, Wartman SA. Perceptions of medical school deans and state medical society executives about physician workforce. JAMA 2003;290:2992–5
10. The Center for Health Workforce Studies. Physician Supply and Demand Indicators in New York, 2000–2005: A Summary of Trends for 35 Specialties. State University of New York, 2006
11. Kheterpal S, Tremper K, Shanks A, Morris M. Seventh and eighth year follow-up on workforce and finances of the United States anesthesiology training programs: 2007 and 2008. Anesth Analg 2009;109:897–9
13. Merwin E, Stern S, Jordan L. Supply and demand of nurse anesthetists: present and future projections. Unpublished information courtesy of Lorraine M. Jordan, PhD, Director of Research and AANA Foundation, 2003
14. Schubert A. National Resident Matching Program results for 2007: recruitment maintains all-time high for specialty of anesthesiology. ASA Newslett 2007;71:17–27
15. Grogono AW. National Resident Matching Program results for 2004: slight decline in recruitment. ASA Newslett 2004;68:18–23
16. Tremper K, Shanks A, Morris M. Five-year follow-up on the work force and finances of United States anesthesiology training programs: 2000 to 2005. Anesth Analg 2007;104:863–68
17. Eckhout G. Where are those anesthesiologists? Deciphering the numbers. ASA Newslett 2004;68:14–6
18. Schubert A, Christiansen EK, Peterson MD. Workforce Contribution of Anesthesiologists over 50: Results of a 2006 National Survey. Proceedings of the ASA Annual Meeting, 2009
19. Ryan-Maher MB. Anesthesiology pay lagging productivity, inflation. Anesthesiol News 2007;33:1, 10
20. ASA House of Delegates Report, 2006
21. Daugherty L, Finseca R, Kumar KB, Michaud PC. An Analysis of the Labor Markets for Anesthesiology. Santa Monica, CA: RAND Corporation, 2010:1–155
22. Fordyce MA, Chen FM, Doescher MP, Hart LG. 2005 Physician Supply and Distribution in Rural Areas of the United States. Rural Health Research and Policy Centers, University of Washington School of Medicine, 2007
23. United States Government Accountability Office. Many Medical Facilities Have Challenges in Recruiting and Retaining Nurse Anesthetists. GAO-08-56, 2007. Available at: http://www.gao.gov/new.items/d0856.pdf
. Accessed July 1, 2011
24. Merwin E, Stern S, Jordan LM, Bucci M. New estimates for CRNA vacancies. AANA J 2009;77:121–9
25. Assembly of School Faculty, , 2007, Preliminary Report. AANA Foundation Faculty Workforce Study
26. Dutton RP. Anesthesia in the United States. Anesthesia Quality Institute, American Society of Anesthesiologists, 2009:1–30
29. Abouleish A, Dexter F, Epstein R, Lubarsky D, Whitten C, Prough D. Labor costs incurred by anesthesiology groups because of operating rooms not being allocated and cases not being scheduled to maximize operating room efficiency. Anesth Analg 2003;96:1109–13
30. Dexter F, Epstein R, Marsh H. A statistical analysis of weekday operating room anesthesia group staffing costs at nine independently managed surgical suites. Anesth Analg 2001;92:1493–8
31. Strum D, Vargas L, May J. Surgical subspecialty block utilization and capacity planning. Anesthesiology 1999;90:1176–85
32. Poley S, Belsky D, Gaul K, Ricketts T, Fraher E, Sheldon G. Longitudinal trends in the U.S. surgical workforce 1981–2006. Bull Am Coll Surg 2009;94:27–31
33. Donen N, King F, Reid D, Blackstock D. Canadian anesthesia physician resources: 1996 and beyond. Can J Anaesth 1999;46:962–9
34. Kaplan RF, Yang CI. Sedation and analgesia in pediatric patients for procedures outside the operating room. Anesthesiol Clin North Am 2002;20:181–94
35. Lalwani K, Michel M. Pediatric sedation in North American children's hospitals: a survey of anesthesia providers. Paediatr Anaesth 2005;15:209–13
36. Berkowitz R, Wadden TA, Tershakovec AM, Cronquist JL. Behavior therapy and sibutramine for the treatment of adolescent obesity: a randomized controlled trial. JAMA 2003;289:1805–12
37. Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, Van Lente F, Levey AS. Prevalence of chronic kidney disease in the United States. JAMA 2007;298:2038–47
38. Centers for Disease Control and Prevention (CDC). Prevalence of self-reported arthritis or chronic joint symptoms among adults: United States, 2001. MMWR Morb Mortal Wkly Rep 2002;51:948–50
39. Boyle JP, Honeycutt AA, Narayan KM, Hoerger TJ, Geiss LS, Chen H, Thompson TJ. Projection of diabetes burden through 2050: impact of changing demography and disease prevalence in the U.S. Diabetes Care 2001;24:1936–40
40. Angus DC, Kelley MA, Schmitz RJ, White A, Popovich J. Current and projected workforce requirements for care of the critically ill and patients with pulmonary disease. JAMA 2000;284:2762–70
41. Etzioni DA, Liu JH, Maggard MA, Ko CY. The aging population and its impact on the surgery workforce. Ann Surg 2003;238:170–7
42. ASA at a glance. ASA Newslett 2008;72:26–7
43. Smart DR, Sellers J. Physician Characteristics and Distribution in the US, 2010 Edition. Chicago: American Medical Association, 2010:21, 439, Tables 1.9 and 6.2
44. Blough GG, Scott SJ. Creative Scheduling for Anesthesiologists: Accommodating the Workforce in 2007, Syllabus of the 2007 ASA Conference on Practice Management
45. Physician Specialty Data: A Chart Book. Washington, DC: Association of American Medical Colleges, 2006
46. Graduate Medical Education in the United States 2005: JAMA 2005;294:1129–32
47. Smart DR, Sellers J. Physician Characteristics and Distribution in the US, 2008 Edition. Chicago: American Medical Association, 2007:20
48. Amburgey B, Fordham M, Payne B, Trebelhorn M. Kentucky State Legislature Study. In: ASA House of Delegates 2007 Session Handbook, 2007:1
49. American Board of Physician Specialties. Available at: www.abpsus.org
. Accessed July 5, 2011
50. Schubert A. National resident matching program results 2010: has recruitment into our specialty crested? ASA Newslett 2010;74:24–9
51. Nakata DA, Powell PT, Choi A. What effect does income have on anesthesiologist retirement rates? Anesthesiology 2004;101:A1325
52. Dorsey ER, Jarjoura D, Rutecki GW. Influence of controllable lifestyle on recent trends in specialty choice by US medical students. JAMA 2003;290:1173–8
53. Dulisse B, Cromwell J. No harm found when nurse anesthetists work without supervision by physicians. Health Aff (Millwood) 2010;29:1469–75
54. Abenstein JP, Long KH, McGlinch BP, Dietz NM. Is physician anesthesia cost-effective? Anesth Analg 2004;98:750–7
55. O'Riordan J, Kozakowski M, Hennessey S, Hosmer K. Value at the Center: State of Our Industry 2007. Washington, DC: The Healthcare Advisory Company, 2007:3
56. National Center for Health Statistics: Ambulatory and Inpatient Procedures [Internet]. U.S. Department of Health and Human Services Centers for Disease Control and Prevention, 2006
57. Williams TE, Satiani B, Thomas A, Ellison EC. The impending shortage and the estimated cost of training the future surgical workforce. Ann Surg 2009;250:590–7
58. Hawkins JL, Gibbs CP, Orleans M, Martin-Salvaj G, Beaty B. Obstetric anesthesia work force survey, 1981 versus 1992. Anesthesiology 1997;87:135–43
59. Health Resources and Services Administration Report to Congress: The Critical Care Workforce—A Study of Supply and Demand for Critical Care Physicians. Washington, DC: U.S. Department of Health and Human Services, 2006:1–34
60. NewsRX™: World's Largest Source of Health Information [Internet]. 2004. Available at: http://www.NewsRx.net
. Accessed July 5, 2011
61. Etzioni DA, Liu JH, Maggard MA, O'Connell JB, Ko CY. Workload projections for surgical oncology: will we need more surgeons? Ann Surg Oncol 2003;10:1112–7
62. Pennington DG. Impact of new technology on cardiothoracic surgical practice. Ann Thorac Surg 2006;81:10–8
63. Freund PR, Posner KL. Sustained increases in productivity with maintenance of quality in an academic anesthesia practice. Anesth Analg 2003;96:1104–8
64. Pondrom S. Can generics be trusted? Am J Transplant 2010;10:708–9
65. Kheterpal S, Tremper K, Shanks A, Morris M. Six-year follow-up on workforce and finances of the United States anesthesiology training programs: 2000 to 2006. Anesth Analg 2009;108:263–72
66. Smith M, Sandberg WS, Foss J, Massoli K, Kanda M, Barsoum W, Schubert A. A high-throughput operating room process for joint arthroplasties routinely outperforms normal counterparts. Anesthesiology 2008;109:25–35
67. Kheterpal S, Tremper KK, Shanks A, Morris M. Brief report: workforce and finances of the United States anesthesiology training programs: 2009–2010. Anesth Analg 2011;112:1480–6
68. Adamy J.. Law Won't Dent Health Outlays. The Wall Street Journal, , 2010
69. Sandberg WS. Barbarians at the gate. Anesth Analg 2009;109:695–9
71. Abouleish AE, Stead SS, Cohen NA. Myth or fact? Nurse anesthetists cost less than anesthesiologists. ASA Newslett 2010;74:30–4
72. The American Academy of Anesthesiologist Assistants. Available at: www.anesthetist.org
. Accessed July 5, 2011
73. Cost Analysis for Anesthesia Practices, 2006: Report Based on 2005 Data. Physician Compensation and Production Survey 2006. Englewood, CO: Medical Group Management Association
79. Gunn IP. Rural health care and the nurse anesthetist. CRNA 2000;11:77–86
80. Silber JH, Kennedy SK, Even-Shoshan O, Chen W, Koziol LF, Showan AM, Longnecker DE. Anesthesiologist direction and patient outcomes. Anesthesiology 2000;93:152–63
81. Record JC, McCally M, Schweitzer SO, Blomquist RM, Berger BD. New health professions after a decade and a half: delegation, productivity and costs in primary care. J Health Polit Policy Law 1980;5:470–97
Supplemental Digital Content
© 2012 International Anesthesia Research Society