Advertisement

The impact of endovascular repair on specialties performing abdominal aortic aneurysm repair

Open ArchivePublished:May 04, 2015DOI:https://doi.org/10.1016/j.jvs.2015.03.042

      Background

      Abdominal aortic aneurysm (AAA) repair has been performed by various surgical specialties for many years. Endovascular aneurysm repair (EVAR) may be a disruptive technology, having an impact on which specialties care for patients with AAA. Therefore, we examined the proportion of AAA repairs performed by various specialties over time in the United States and evaluated the impact of the introduction of EVAR.

      Methods

      The Nationwide Inpatient Sample (2001-2009) was queried for intact and ruptured AAA and for open repair and EVAR. Specific procedures were used to identify vascular surgeons (VSs), cardiac surgeons (CSs), and general surgeons (GSs) as well as interventional cardiologists and interventional radiologists for states that reported unique treating physician identifiers. Annual procedure volumes were subsequently calculated for each specialty.

      Results

      We identified 108,587 EVARs and 85,080 open AAA repairs (3011 EVARs and 12,811 open repairs for ruptured AAA). VSs performed an increasing proportion of AAA repairs during the study period (52% in 2001 to 66% in 2009; P < .001). GSs and CSs performed fewer repairs during the same period (25% to 17% [P < .001] and 19% to 13% [P < .001], respectively). EVAR was increasingly used for intact (33% to 78% of annual cases; P < .001) as well as ruptured AAA repair (5% to 28%; P < .001). The proportion of intact open repairs performed by VSs increased from 52% to 65% (P < .001), whereas for EVAR, the proportion went from 60% to 67% (P < .001). The proportion performed by VSs increased for ruptured open repairs from 37% to 53% (P < .001) and for ruptured EVARs from 28% to 73% (P < .001). Compared with treatment by VSs, treatment by a CS (0.55 [0.53-0.56]) and GS (0.66 [0.64-0.68]) was associated with a decreased likelihood of undergoing endovascular rather than open AAA repair.

      Conclusions

      VSs are performing an increasing majority of AAA repairs, in large part driven by the increased utilization of EVAR for both intact and ruptured AAA repair. However, GSs and CSs still perform AAA repair. Further studies should examine the implications of these national trends on the outcome of AAA repair.
      During the late 20th century, surgery became a technology-driven profession.
      • Riskin D.J.
      • Longaker M.T.
      • Gertner M.
      • Krummel T.M.
      Innovation in surgery: a historical perspective.
      Since then, innovations such as endoscopic and endovascular surgery have transformed clinical medicine. Besides changing the procedure itself, these disruptive technologies have had their effect on the type of physicians performing the procedures. Percutaneous coronary intervention, for example, has diminished the proportion of coronary revascularizations performed by cardiac surgeons (CSs), whereas the proportion of interventional cardiologists (ICs) increased dramatically with the use of this technique.
      • Epstein A.J.
      • Polsky D.
      • Yang F.
      • Yang L.
      • Groeneveld P.W.
      Coronary revascularization trends in the United States, 2001-2008.
      For abdominal aortic aneurysm (AAA) repair, it is unclear how the introduction and widespread adoption of endovascular aneurysm repair (EVAR) have changed the distribution of specialties performing elective and ruptured AAA repair.
      Before the introduction of EVAR, open surgical repair was the primary method of treatment. Using Medicare data, Birkmeyer et al
      • Birkmeyer J.D.
      • Stukel T.A.
      • Siewers A.E.
      • Goodney P.P.
      • Wennberg D.E.
      • Lucas F.L.
      Surgeon volume and operative mortality in the United States.
      showed that between 1998 and 1999, before the widespread adoption of EVAR, vascular surgeons (VSs) performed 39% of all elective AAA repairs, whereas CSs and general surgeons (GSs) performed 33% and 28%, respectively. In contrast to elective AAA repair, GSs performed the largest proportion of ruptured AAA repairs at 39%, followed by VSs at 33% and CSs at 29%.
      Currently, as with coronary revascularization, the endovascular approach has also led to the inclusion of nonsurgical specialists in treating patients with AAA, such as ICs and interventional radiologists (IRs). Because the performance of EVAR requires a specific skill set that has not been mastered by many surgeons from other specialties, we hypothesize that the proportion of surgical specialists other than VSs (ie, GSs and CSs) has declined, whereas VSs, IRs, and ICs are responsible for an increasing number of patients because of a shift from open repair toward EVAR.
      The purpose of this study was to analyze how the introduction of EVAR has influenced which specialties are providing care for AAA patients for both elective and ruptured AAA repair in the United States.

      Methods

       Database

      The Nationwide Inpatient Sample (NIS) is the largest national administrative database and represents a 20% sample of all payer (insured and uninsured) hospitalizations. The NIS is maintained by the Agency for Healthcare Research and Quality as part of the Healthcare Cost and Utilization Project. Years 2001 to 2009 were queried using International Classification of Diseases, Ninth Revision (ICD-9) codes to identify patients with diagnosis codes for intact (ie, elective, symptomatic, and mycotic aneurysms) AAA (441.4) and ruptured AAA (441.3). ICD-9 coding does not distinguish infrarenal from juxtarenal or suprarenal AAA. More recent years could not be interrogated because of discontinuation of the surgeon identification variables in the NIS database after 2009.

      Availability of data elements in the 1988-2012 National Inpatient Sample (NIS) [press release]. NIS Database Documentation: HCUP2013.

      Patients who underwent open AAA repair (38.44, 39.25) or EVAR (39.71) were selected. Patients with procedural codes for both open repair and EVAR were considered to have undergone EVAR as they likely represent conversions to open repair. Patients with ICD-9 codes for a thoracic aneurysm (441.1 or 441.2), thoracoabdominal aneurysm (441.6 or 441.7), or aortic dissection (441.00-441.03) were excluded. As the NIS contains de-identified data only without protected health information, Institutional Review Board approval and patient consent were waived.
      The primary outcome was proportional procedure volume by physician specialty over time for intact and ruptured AAA repair. We evaluated the uptake of EVAR overall and by specialty over time. In addition, we assessed the likelihood of receiving EVAR rather than open repair by specialty.

       Physician specialty

      For AAA repair, we were interested in the following types of physicians: VSs, GSs, CSs, ICs, and IRs. The NIS provides unique physician identifiers per state that allow tracking of procedures performed by that physician during that specific year in that state. Of the available states, 27 provide two unique physician identifiers, with 22 of the 27 specifically detailing which physician performed the primary procedure (Supplementary Table I, online only). For the remaining five states, the identifiers were used only when both identifiers were the same to ensure that the identified physician was the one performing the primary procedure. We composed a list of specific procedures (Supplementary Table II, online only) that we used to determine the specialty of each physician (VS, GS, CS, IC, or IR). The top 15 procedures identified for each of the physician specialties are listed in Supplementary Table III, online only. Similar approaches have been previously reported.
      • Csikesz N.G.
      • Simons J.P.
      • Tseng J.F.
      • Shah S.A.
      Surgical specialization and operative mortality in hepato-pancreatico-biliary (HPB) surgery.
      • Schipper P.H.
      • Diggs B.S.
      • Ungerleider R.M.
      • Welke K.F.
      The influence of surgeon specialty on outcomes in general thoracic surgery: a national sample 1996 to 2005.
      • Vogel T.R.
      • Dombrovskiy V.Y.
      • Carson J.L.
      • Haser P.B.
      • Graham A.M.
      Lower extremity angioplasty: impact of practitioner specialty and volume on practice patterns and healthcare resource utilization.
      Subsequently, a hierarchical model was created: each physician who performed >10 cardiac surgery procedures was labeled a CS; the remaining physicians who performed >10 interventional cardiology procedures (eg, coronary stenting) were labeled ICs; physicians with >10 interventional radiology procedures not typically performed by VSs (eg, liver biopsy, nephrostomy) were identified as IRs; the remaining physicians whose procedures consisted of 75% to 100% of vascular procedures with >10 in number were classified as VSs; physicians whose procedures consisted of 0% to 75% of vascular procedures and performed >10 general surgery procedures were classified as GSs. Similar approaches have been previously described.
      • Dimick J.B.
      • Cowan Jr., J.A.
      • Stanley J.C.
      • Henke P.K.
      • Pronovost P.J.
      • Upchurch Jr., G.R.
      Surgeon specialty and provider volumes are related to outcome of intact abdominal aortic aneurysm repair in the United States.
      • Liang P.
      • Hurks R.
      • Bensley R.P.
      • Hamdan A.
      • Wyers M.
      • Chaikof E.
      • et al.
      The rise and fall of renal artery angioplasty and stenting in the United States, 1988-2009.
      Of the procedures labeled as open repairs, 210 were coded as being performed by ICs or IRs (0.1% of total procedures). We thought these were most likely miscoded endovascular procedures and excluded these patients from further analyses.

       Statistical approach

      Mean and standard deviation are reported for parametric data. Baseline variables were compared by χ2 tests or t-tests, where appropriate. We examined the proportional volume of open AAA repairs and EVAR for each specialty and how this changed during the study period. Trends over time were assessed by the Cochran-Armitage test for trend. A P value < .05 indicates that annual procedural volumes followed a significant upward or downward (ie, nonrandom) trend over time. Multivariable logistic regression analysis was conducted to examine the influence of physician specialty type on the type of procedure performed, whether open or endovascular. Analyses were considered statistically significant when P < .05. Statistical analyses were performed with SAS 9.2 software (SAS Institute, Cary, NC) and SPSS Statistics 21 (IBM Inc, Chicago, Ill).

      Results

      Overall, 108,587 EVARs and 85,080 open AAA repairs were identified in the study period, of which 3011 EVARs and 12,811 open repairs were for ruptured AAA. The annual overall volume increased from 20,134 in 2001 to 22,541 in 2009 (P < .001). Characteristics of the patients and hospitals are detailed in Table I. Of all AAA repairs, 61% of AAA repairs were performed by VSs, 20% by GSs, and 16% by CSs, whereas the remainder were performed by ICs and IRs (3% combined). Fig 1 illustrates changes over time for each physician specialty. VSs performed an increasing proportion of AAA repairs during the study period (52% in 2001 to 66% in 2009; P < .001; Supplementary Table IV, online only). During the same period, GSs and CSs performed fewer repairs (25% to 17% [P < .001] and 19% to 13% [P < .001], respectively). Similarly, the absolute number of VSs performing AAA repair increased 30% during the study period, whereas the number of GSs and CSs decreased over time (46% and 30%, respectively).
      Table IDemographic and comorbidity characteristics of patients undergoing open aortic aneurysm repair or endovascular aneurysm repair (EVAR) per physician specialty
      Open repairEVAROpen EVAR
      VSsGSsCSsP valueVSsGSsCSsICsIRsP valueOverallP value
      No.45,80421,62517,65172,48917,50014,0333055151085,080108,587
      Age, years71.371.971<.00173.774.072.873.174.1<.00171.473.6<.001
      Female25.523.623.0<.00118.316.915.416.817.0<.00124.517.7<.001
      White race90.490.791.0.05191.591.292.287.378.7<.00190.691.2<.001
      Teaching hospital60.530.745.1<.00163.933.848.446.341.0<.00149.756.2<.001
      Urban location95.588.293.6<.00195.492.893.910097.2<.00193.295.0<.001
      Hospital bed size<.001<.001<.001
       Small7.08.25.58.64.48.213.00.07.07.9
       Medium17.926.723.015.522.119.59.424.421.217.0
       Large75.165.171.575.973.572.377.675.671.875.1
      Emergent admission13.621.011.5<.0013.02.62.01.43.0<.00115.12.8<.001
      CSs, Cardiac surgeons; GSs, general surgeons; ICs, interventional cardiologists; IRs, interventional radiologists; VSs, vascular surgeons.
      Categorical variables are presented as percentage frequencies.
      P values compare differences within the treatment groups.
      Figure thumbnail gr1
      Fig 1Proportion of all abdominal aortic aneurysm (AAA) repairs (both open and endovascular) performed by each physician specialty from 2001 to 2009 in the Nationwide Inpatient Sample (NIS; totals sum to 100%). CS, Cardiac surgeon; GS, general surgeon; IC, interventional cardiologist; IR, interventional radiologist; VS, vascular surgeon.

       Intact AAA repair

      With 55%, VSs performed the majority of open AAA repairs (increasing from 52% to 65% from 2001 to 2009; P < .001). During this same period, GSs performed 24% of all intact open repairs (decreasing from 25% to 16%; P < .001), followed by CSs with 22% of cases (24% to 19%; P < .001; Fig 2, A). VSs also performed the majority of EVARs at 67% (increasing from 60% to 67% from 2001 to 2009), followed by 16% performed by GSs (19% to 17%; P < .001), 13% by CSs (10.5% to 11.3%; P = .009), and 4% by ICs and IRs combined (10% to 6%; P = .015; Fig 2, B). The absolute number of EVARs increased from 5906 in 2001 (33% of the annual intact AAA repairs) to 16,252 in 2009 (78%). During the same period, the number of open repairs decreased from 12,188 (67%) to 4678 (22%). Consequently, EVAR has become the primary treatment method for intact AAA in all three surgical specialties (Fig 3). Because VSs perform a greater proportion of endovascular procedures, the rise in EVAR utilization has in part led to VSs performing an increasing majority of overall intact AAA repairs (54% in 2001 to 66% in 2009; P < .001).
      Figure thumbnail gr2
      Fig 2A, Proportion of all open repairs for intact abdominal aortic aneurysms (AAAs) performed by physician specialty from 2001 to 2009 in the Nationwide Inpatient Sample (NIS; totals sum to 100%). B, Proportion of all endovascular repairs for intact AAAs performed by physician specialty from 2001 to 2009 in the NIS (totals sum to 100%). CS, Cardiac surgeon; GS, general surgeon; IC, interventional cardiologist; IR, interventional radiologist; VS, vascular surgeon.
      Figure thumbnail gr3
      Fig 3Proportion of intact abdominal aortic aneurysms (AAAs) treated by endovascular repair within each specialty from 2001 to 2009 in the Nationwide Inpatient Sample (NIS). CS, Cardiac surgeon; GS, general surgeon; VS, vascular surgeon.

       Ruptured AAA repair

      VSs performed 49% of open ruptured AAA repairs (increasing from 37% to 53%; P < .001; Fig 4, A), followed by GSs with 35% (44% to 33%; P < .001) and CSs with 16% (19% to 14%; P = .001). With 73%, VSs also carried out the majority of ruptured EVARs (28% to 73%; P < .001; Fig 4, B), whereas GSs performed 15% (54% to 16%; P < .001) and CSs 9% (18% to 8%; P = .008). ICs and IRs together are responsible for 3% of ruptured EVARs (0% to 3%; P = .095). A dramatic overall increase in the utilization of ruptured EVAR was observed (5% to 38% of the annual ruptured volume). This was most pronounced for VSs, for whom the utilization of ruptured EVAR went from 4% in 2001 to 46% in 2009 (P < .001; Fig 5).
      Figure thumbnail gr4
      Fig 4A, Proportion of all open ruptured abdominal aortic aneurysm (AAA) repairs performed by physician specialty from 2001 to 2009 in the Nationwide Inpatient Sample (NIS; totals sum to 100%). B, Proportion of all endovascular aneurysm repairs (EVARs) for ruptured AAAs performed by physician specialty from 2001 to 2009 in the NIS (totals sum to 100%). CS, Cardiac surgeon; GS, general surgeon; IC, interventional cardiologist; IR, interventional radiologist; VS, vascular surgeon.
      Figure thumbnail gr5
      Fig 5Proportion of ruptured abdominal aortic aneurysms (AAAs) treated by endovascular repair within each specialty from 2001 to 2009 in the Nationwide Inpatient Sample (NIS). CS, Cardiac surgeon; GS, general surgeon; VS, vascular surgeon.

       Likelihood of receiving EVAR

      Compared with treatment by VSs, treatment by CSs (odds ratio [OR], 0.55; 95% confidence interval [CI], 0.53-0.56) and GSs (OR, 0.66; 95% CI, 0.64-0.68) was associated with a significantly lower likelihood of receiving EVAR (Table II). In addition, women (OR, 0.57; 95% CI, 0.55-0.58), those with nonwhite race (OR, 0.88; 95% CI, 0.84-0.91), and emergency admission (OR, 0.14; 95% CI, 0.14-0.15) were significantly less likely to undergo EVAR. Advanced age (OR, 1.42; 95% CI, 1.40-1.44; per 10 years), treatment in a teaching hospital (OR, 1.27; 95% CI, 1.25-1.30), and urban designation of the hospital (OR, 1.33; 95% CI, 1.27-1.40) were predictive of EVAR. During the study period, the probability for receiving EVAR increased annually (OR, 1.33; 95% CI, 1.32-1.34; per year).
      Table IIMultivariable predictors for the likelihood of receiving endovascular aneurysm repair (EVAR)
      VariableOR95% CIP value
      Surgeon specialty
       VSsReference
       CSs0.550.53-0.56<.001
       GSs0.660.64-0.68<.001
      Emergent admission0.140.14-0.15<.001
      Age (per 10 years)1.421.40-1.44<.001
      Female sex0.570.55-0.58<.001
      Nonwhite race0.880.84-0.91<.001
      Year of surgery (per year)1.331.32-1.34<.001
      Teaching hospital1.271.25-1.30<.001
      Hospital location1.331.27-1.40<.001
      Hospital bed size
       SmallReference
       Medium0.720.69-0.76<.001
       Large0.980.94-1.02<.319
      CI, Confidence interval; CSs, cardiac surgeons; GSs, general surgeons; OR, odds ratio; VSs, vascular surgeons.
      OR >1 predicts EVAR.

      Discussion

      The main finding of this study is that with the introduction of EVAR, the proportion of AAA repairs being performed by each physician specialty has changed. VSs performed an increasing majority of both open and endovascular intact AAA repairs, whereas the proportion carried out by CSs and GSs has steadily declined. The distribution of specialties performing ruptured AAA repair shifted from predominantly GSs in the first years of the study toward VSs in later years. During the study period, EVAR has become the dominant treatment for intact AAAs and is being used for an increasing number of ruptured AAAs as well. As EVAR is most likely performed by VSs, the overall proportion of repairs done by VSs—intact and ruptured—increased substantially with the widespread adoption of EVAR.
      Regarding intact AAA repair, our results are in line with a study by Birkmeyer et al, showing that between 1998 and 1999, intact open repairs were predominantly performed by VSs. However, Birkmeyer et al found a relatively even distribution, with 39% being done by VSs, 33% by CSs, and 28% by GSs, whereas our results showed that VSs already performed a majority of the open repairs in the early years of the study, and this difference continued to increase over time. Regionalization of open AAA repairs to high-volume centers during the turn of the century is likely to have contributed to VSs being increasingly responsible for AAA surgery.
      • Hill J.S.
      • McPhee J.T.
      • Messina L.M.
      • Ciocca R.G.
      • Eslami M.H.
      Regionalization of abdominal aortic aneurysm repair: evidence of a shift to high-volume centers in the endovascular era.
      In addition, retirement of senior surgeons who were trained at a time when GSs and CSs customarily performed repair of AAAs and may have been less likely to obtain endovascular skills may have added to the shift toward VS. Similar to the early years of our study, previous reports show that GSs performed the majority of ruptured AAA repairs at 39%, followed by VSs at 33% and CSs at 29%, before the introduction of EVAR.
      • Cronenwett J.L.
      • Birkmeyer J.D.
      The Dartmouth Atlas of Vascular Health Care.
      As EVAR became more widely used in the emergency setting, this distribution changed toward a growing proportion of emergency repairs being performed by VSs. Because VSs also performed an increasing proportion of open ruptured AAA repairs separately from trends in endovascular repair, centralization of ruptured AAA care is likely to have contributed to the shift from GSs toward VSs as well. Yet a persistent presence of GSs treating open ruptured AAAs remained, which could be due to geographic location where the presence of VSs may be lacking.
      • Maybury R.S.
      • Chang D.C.
      • Freischlag J.A.
      Rural hospitals face a higher burden of ruptured abdominal aortic aneurysm and are more likely to transfer patients for emergent repair.
      In these areas, the emergent nature of a ruptured AAA may preclude the transfer of the patient to a center with VSs, necessitating immediate treatment by an available GS.
      The phenomenon of disruptive technologies in health care is not new. In coronary artery disease, the number of coronary revascularizations performed with coronary stenting rapidly increased after the first coronary stent was introduced in 1994, whereas the utilization of coronary bypass grafting declined.
      • Epstein A.J.
      • Polsky D.
      • Yang F.
      • Yang L.
      • Groeneveld P.W.
      Coronary revascularization trends in the United States, 2001-2008.
      As a result, ICs rather than CSs currently perform the majority of coronary revascularizations. A similar shift was seen in vascular surgery with the introduction of carotid stenting. Before its introduction, carotid revascularization through endarterectomy was predominantly done by VSs and, to a lesser extent, GSs, CSs, and neurosurgeons.
      • Hollenbeak C.S.
      • Bowman A.R.
      • Harbaugh R.E.
      • Casale P.N.
      • Han D.
      The impact of surgical specialty on outcomes for carotid endarterectomy.
      We noted that GS and CS practice included a substantial percentage of carotid endarterectomies (8.6% and 10.0% of the selected procedures we identified, respectively; Supplementary Table III, online only). However, we did not evaluate changes in carotid revascularization over time in this study. After Food and Drug Administration approval in 2004, carotid stenting is increasingly used with rapid adoption by not only surgeons but also IRs and ICs.
      • Rosenfield K.
      • Babb J.D.
      • Cates C.U.
      • Cowley M.J.
      • Feldman T.
      • Gallagher A.
      • et al.
      Clinical competence statement on carotid stenting: training and credentialing for carotid stenting—multispecialty consensus recommendations: a report of the SCAI/SVMB/SVS Writing Committee to develop a clinical competence statement on carotid interventions.
      • Gray W.A.
      A cardiologist in the carotids.
      Currently, carotid endarterectomy use is still declining, whereas some patients are being treated through stenting predominantly by ICs.
      • Nallamothu B.K.
      • Lu M.
      • Rogers M.A.
      • Gurm H.S.
      • Birkmeyer J.D.
      Physician specialty and carotid stenting among elderly medicare beneficiaries in the United States.
      • Skerritt M.R.
      • Block R.C.
      • Pearson T.A.
      • Young K.C.
      Carotid endarterectomy and carotid artery stenting utilization trends over time.
      EVAR has certainly changed how AAA is being treated, but contrary to the examples given before, VSs have only increased their role as the treating surgeons for intact and ruptured AAAs.
      Our study has several limitations that should be addressed. Because administrative databases were used, important clinical data, such as anatomic information or hemodynamic status, which could influence the choice of procedure and subsequent outcomes, could not be assessed. In addition, the difficulty of distinguishing a pre-existing comorbid condition from a postoperative complication in this data set makes an adequate risk adjustment model difficult. Therefore, we chose not to perform outcomes analysis using the NIS. However, the NIS does afford national representation of all age groups, thus making it an optimal source of epidemiologic data. Also, as the NIS database does not include the specialty of the attending physician, we employed an algorithm incorporating specialty-specific procedures similar to what has been described before.
      • Schipper P.H.
      • Diggs B.S.
      • Ungerleider R.M.
      • Welke K.F.
      The influence of surgeon specialty on outcomes in general thoracic surgery: a national sample 1996 to 2005.
      • Eppsteiner R.W.
      • Csikesz N.G.
      • Simons J.P.
      • Tseng J.F.
      • Shah S.A.
      High volume and outcome after liver resection: surgeon or center?.
      • Eslami M.H.
      • Csikesz N.
      • Schanzer A.
      • Messina L.M.
      Peripheral arterial interventions: trends in market share and outcomes by specialty, 1998-2005.
      Our algorithm for identifying VSs is arbitrary (75% vascular surgery cases), suggesting that some board-certified VSs may have been mistakenly classified as GSs or vice versa. However, our methodology reflects what physicians were actually doing routinely in practice and focused on the change over time. In addition, our algorithm generated similar distributions of physician's specialty performing open repair in the early years of the current studied period compared with a previously published large study using physician self-identification of specialty.
      Further, procedures coded as open repairs performed by ICs or IRs (0.1%) were excluded from this study, as they are, most likely, miscoded endovascular procedures. Unfortunately, it is not possible to identify similar procedural coding errors for the remaining open repairs. Consequently, a very small proportion of miscoded procedures may have remained. Finally, the discontinuation of the surgeon-identifying variable after 2009 prevented the inclusion of more recent years. However, as is demonstrated in this study, the major shift from open repair to EVAR was already well established by 2009.
      • Dua A.
      • Kuy S.
      • Lee C.J.
      • Upchurch Jr., G.R.
      • Desai S.S.
      Epidemiology of aortic aneurysm repair in the United States from 2000 to 2010.

      Conclusions

      Our results show that VSs are performing an increasing majority of AAA repairs, in large part driven by the increased utilization of EVAR for both intact and ruptured AAA repair. However, GSs and CSs still perform AAA repair. Treatment by GSs and CSs, emergent admission, female sex, and nonwhite race are associated with a decreased likelihood of receiving EVAR. Advanced age, more recent year of surgery, treatment in a teaching hospital, and urban designated area of the hospital increased the probability of receiving EVAR. Further studies should examine the implications of these national trends on the outcome of AAA repair.

      Author contributions

      Conception and design: RH, KU, MS
      Analysis and interpretation: RH, KU, GD, MS
      Data collection: Not applicable
      Writing the article: RH, KU, DB, PS
      Critical revision of the article: DB, JH, HV, MS
      Final approval of the article: MS
      Statistical analysis: RH, KU, GD
      Obtained funding: Not applicable
      Overall responsibility: MS
      KU and RH contributed equally to this article and share co-first authorship.

      Appendix.

      Supplementary Table I (online only)Proportion of the cohort per state
      StateProportion of cohort, %
      Arkansas1.1
      Arizona5.3
      Colorado2.0
      Florida17.6
      Georgia0.4
      Iowa2.1
      Kansas0.7
      Kentucky3.2
      Maryland4.6
      Maine0.1
      Michigan4.2
      Minnesota0.9
      Missouri5.9
      Montana0.1
      Nebraska0.9
      New Hampshire2.0
      New Jersey5.8
      Nevada0.8
      New York14.4
      Oregon0.6
      Pennsylvania5.8
      Rhode Island0.2
      South Carolina1.4
      South Dakota0.1
      Tennessee5.4
      Texas8.9
      Virginia4.3
      Washington1.4
      West Virginia<0.1
      Wyoming<0.1
      Supplementary Table II (online only)Procedures used to identify physician specialty
      ICD-9Description
      VS
       38.12Carotid endarterectomy
       39.29Peripheral vascular bypass
       84.15Below-knee amputation
       84.17Above-knee amputation
      GS
       17.11-24; 53.00-9Hernia repair
       47.01-19Appendectomy
       51.21-24Cholecystectomy
      CS
       36.10-19Coronary artery bypass grafting
       35.20-28Heart valve replacement
       39.61Cardiopulmonary bypass
      IC
       00.66; 36.01-02; 36.05Percutaneous transmural coronary angioplasty
       36.04Intracoronary thrombolysis
       36.06-07Intracoronary stenting
       37.21-23Heart catheterization
      IR
       33.26Closed lung biopsy
       39.1Transjugular intrahepatic portosystemic shunt
       50.11Closed liver biopsy
       55.03-04Percutaneous nephrostomy
       78.49; 81.65Percutaneous vertebroplasty
       99.25Chemoembolization
      CS, Cardiac surgeon; GS, general surgeon; IC, interventional cardiologist; ICD-9, International Classification of Diseases, Ninth Revision; IR, interventional radiologist; VS, vascular surgeon.
      Supplementary Table III (online only)Top 15 procedures performed per physician specialty
      ICD-9%Description
      VS
       38.1217.6Carotid endarterectomy
       39.2910.0Peripheral vascular bypass
       39.57.7Angioplasty or atherectomy of other noncoronary vessel(s)
       39.715.2Endovascular implantation of graft in abdominal aorta
       38.73.4Interruption of the vena cava
       39.493.3Revision of anastomosis of blood vessel or vascular procedure
       39.273.0Arteriovenostomy for renal dialysis
       38.442.8Resection of vessel with replacement, aorta, abdominal
       86.222.5Excisional débridement of wound, infection, or burn
       39.252.4Aorta-iliac-femoral bypass
       84.152.2Other amputation below knee
       84.172.2Amputation above knee
       38.952.1Venous catheterization for renal dialysis
       38.181.8Endarterectomy, lower limb arteries
       84.111.7Amputation of toe
      GS
       51.2310.2Laparoscopic cholecystectomy
       38.128.6Carotid endarterectomy
       39.294.1Peripheral vascular bypass
       47.093.8Other appendectomy
       47.013.1Laparoscopic appendectomy
       45.732.5Open and other right hemicolectomy
       38.932.5Venous catheterization
       86.222.4Excisional débridement of wound, infection, or burn
       45.762.1Open and other sigmoidectomy
       39.52.0Angioplasty or atherectomy of other noncoronary vessel(s)
       51.221.9Cholecystectomy
       39.271.8Arteriovenostomy for renal dialysis
       38.71.7Interruption of vena cava
       86.041.6Other incision with drainage of skin and subcutaneous tissue
       54.591.6Other lysis of peritoneal adhesions
      CS
       38.1210.0Carotid endarterectomy
       36.129.1(Aorto)coronary bypass of two coronary arteries
       36.138.9(Aorto)coronary bypass of three coronary arteries
       36.144.2(Aorto)coronary bypass of four or more coronary arteries
       36.113.5(Aorto)coronary bypass of one coronary artery
       36.153.2Single internal mammary-coronary artery bypass
       39.293.1Peripheral vascular bypass
       35.222.9Other replacement of aortic valve
       35.212.9Replacement of aortic valve with tissue graft
       38.441.7Resection of vessel with replacement, aorta, abdominal
       32.41.7Lobectomy of lung
       39.711.5Endovascular implantation of graft in abdominal aorta
       39.51.3Angioplasty or atherectomy of other noncoronary vessel(s)
       32.291.3Other local excision or destruction of lesion
       35.121.1Open heart valvuloplasty of mitral valve without replacement
      IC
       0.6621.4Percutaneous transmural coronary angioplasty
       37.2213.7Left-sided heart cardiac catheterization
       39.511.7Angioplasty or atherectomy of other noncoronary vessel(s)
       36.0111.6Percutaneous transmural coronary angioplasty
       39.712.5Endovascular implantation of graft in abdominal aorta
       36.052.3Percutaneous transmural coronary angioplasty
       0.612.1Percutaneous angioplasty or atherectomy of precerebral extracranial vessel(s)
       37.232.1Combined right- and left-sided heart cardiac catheterization
       37.721.9Initial insertion of transvenous leads [electrodes] into atrium and ventricle
       37.611.2Implantation of pulsation balloon
       88.721.1Diagnostic ultrasound of heart
       641.0Circumcision
       57.940.9Insertion of indwelling catheter
       35.520.9Repair of atrial septal defect with prosthesis, closed technique
       88.560.7Coronary arteriography using two catheters
      IR
       38.918.3Venous catheterization, not elsewhere classified
       39.511.9Angioplasty or atherectomy of other noncoronary vessel(s)
       38.77.0Interruption of the vena cava
       54.916.8Percutaneous abdominal drainage
       34.916.1Thoracentesis
       38.952.9Venous catheterization for renal dialysis
       55.032.4Percutaneous nephrostomy without fragmentation
       50.112.1Closed (percutaneous) [needle] biopsy of liver
       88.412.1Arteriography of cerebral arteries
       39.712.0Endovascular implantation of graft in abdominal aorta
       33.262.0Closed (percutaneous) [needle] biopsy of lung
       81.661.7Percutaneous vertebral augmentation
       34.041.3Insertion of intercostal catheter for drainage
       88.421.3Aortography
       99.291.2Injection or infusion of other therapeutic or prophylactic substance
      CS, Cardiac surgeon; GS, general surgeon; IC, interventional cardiologist; ICD-9, International Classification of Diseases, Ninth Revision; IR, interventional radiologist; VS, vascular surgeon.
      Supplementary Table IV (online only)Annual proportions (%) of abdominal aortic aneurysm (AAA) repairs performed per specialty
      200120022003200420052006200720082009P value
      Overall
       VSs52.454.356.467.158.665.265.162.265.8<.001
       GSs25.225.424.118.622.516.916.117.917.2<.001
       CSs19.418.917.913.217.316.216.216.112.9<.001
       ICs + IRs3.11.41.71.11.61.62.63.84.0<.001
      Intact open
       VSs51.651.951.658.251.256.959.856.365.1<.001
       GSs24.726.226.922.727.022.818.219.116.0<.001
       CSs23.721.921.519.121.920.422.024.718.9.001
      Intact EVAR
       VSs59.661.663.275.965.170.067.763.966.5<.001
       GSs19.420.118.813.817.212.714.617.016.7<.001
       CSs10.514.214.08.214.614.714.014.011.3.009
       ICs + IRs10.44.14.12.13.12.53.75.15.5.015
      Ruptured open
       VSs36.643.650.059.947.847.952.054.352.7<.001
       GSs44.139.034.027.639.036.728.631.433.3<.001
       CSs19.317.416.012.513.215.419.414.313.9.001
      Ruptured EVAR
       VSs27.540.264.677.075.881.274.779.273.4<.001
       GSs54.133.624.911.012.714.49.96.315.8<.001
       CSs18.322.410.54.99.63.312.910.47.5.008
       ICs + IRs03.807.11.91.22.74.13.4.095
      CSs, Cardiac surgeons; EVAR, endovascular aneurysm repair; GSs, general surgeons; ICs, interventional cardiologists; IRs, interventional radiologists; VSs, vascular surgeons.

      References

        • Riskin D.J.
        • Longaker M.T.
        • Gertner M.
        • Krummel T.M.
        Innovation in surgery: a historical perspective.
        Ann Surg. 2006; 244: 686-693
        • Epstein A.J.
        • Polsky D.
        • Yang F.
        • Yang L.
        • Groeneveld P.W.
        Coronary revascularization trends in the United States, 2001-2008.
        JAMA. 2011; 305: 1769-1776
        • Birkmeyer J.D.
        • Stukel T.A.
        • Siewers A.E.
        • Goodney P.P.
        • Wennberg D.E.
        • Lucas F.L.
        Surgeon volume and operative mortality in the United States.
        N Engl J Med. 2003; 349: 2117-2127
      1. Cronenwett J.L. Birkmeyer J. The Dartmouth atlas of vascular health care. AHA Press, Chicago2000
      2. Availability of data elements in the 1988-2012 National Inpatient Sample (NIS) [press release]. NIS Database Documentation: HCUP2013.

        • Csikesz N.G.
        • Simons J.P.
        • Tseng J.F.
        • Shah S.A.
        Surgical specialization and operative mortality in hepato-pancreatico-biliary (HPB) surgery.
        J Gastrointest Surg. 2008; 12: 1534-1539
        • Schipper P.H.
        • Diggs B.S.
        • Ungerleider R.M.
        • Welke K.F.
        The influence of surgeon specialty on outcomes in general thoracic surgery: a national sample 1996 to 2005.
        Ann Thorac Surg. 2009; 88 (discussion: 1572-3): 1566-1572
        • Vogel T.R.
        • Dombrovskiy V.Y.
        • Carson J.L.
        • Haser P.B.
        • Graham A.M.
        Lower extremity angioplasty: impact of practitioner specialty and volume on practice patterns and healthcare resource utilization.
        J Vasc Surg. 2009; 50 (discussion: 1324-5): 1320-1324
        • Dimick J.B.
        • Cowan Jr., J.A.
        • Stanley J.C.
        • Henke P.K.
        • Pronovost P.J.
        • Upchurch Jr., G.R.
        Surgeon specialty and provider volumes are related to outcome of intact abdominal aortic aneurysm repair in the United States.
        J Vasc Surg. 2003; 38: 739-744
        • Liang P.
        • Hurks R.
        • Bensley R.P.
        • Hamdan A.
        • Wyers M.
        • Chaikof E.
        • et al.
        The rise and fall of renal artery angioplasty and stenting in the United States, 1988-2009.
        J Vasc Surg. 2013; 58: 1331-1338.e1
        • Hill J.S.
        • McPhee J.T.
        • Messina L.M.
        • Ciocca R.G.
        • Eslami M.H.
        Regionalization of abdominal aortic aneurysm repair: evidence of a shift to high-volume centers in the endovascular era.
        J Vasc Surg. 2008; 48: 29-36
        • Cronenwett J.L.
        • Birkmeyer J.D.
        The Dartmouth Atlas of Vascular Health Care.
        Cardiovasc Surg. 2000; 8: 409-410
        • Maybury R.S.
        • Chang D.C.
        • Freischlag J.A.
        Rural hospitals face a higher burden of ruptured abdominal aortic aneurysm and are more likely to transfer patients for emergent repair.
        J Am Coll Surg. 2011; 212: 1061-1067
        • Hollenbeak C.S.
        • Bowman A.R.
        • Harbaugh R.E.
        • Casale P.N.
        • Han D.
        The impact of surgical specialty on outcomes for carotid endarterectomy.
        J Surg Res. 2010; 159: 595-602
        • Rosenfield K.
        • Babb J.D.
        • Cates C.U.
        • Cowley M.J.
        • Feldman T.
        • Gallagher A.
        • et al.
        Clinical competence statement on carotid stenting: training and credentialing for carotid stenting—multispecialty consensus recommendations: a report of the SCAI/SVMB/SVS Writing Committee to develop a clinical competence statement on carotid interventions.
        J Am Coll Cardiol. 2005; 45: 165-174
        • Gray W.A.
        A cardiologist in the carotids.
        J Am Coll Cardiol. 2004; 43: 1602-1605
        • Nallamothu B.K.
        • Lu M.
        • Rogers M.A.
        • Gurm H.S.
        • Birkmeyer J.D.
        Physician specialty and carotid stenting among elderly medicare beneficiaries in the United States.
        Arch Intern Med. 2011; 171: 1804-1810
        • Skerritt M.R.
        • Block R.C.
        • Pearson T.A.
        • Young K.C.
        Carotid endarterectomy and carotid artery stenting utilization trends over time.
        BMC Neurol. 2012; 12: 17
        • Eppsteiner R.W.
        • Csikesz N.G.
        • Simons J.P.
        • Tseng J.F.
        • Shah S.A.
        High volume and outcome after liver resection: surgeon or center?.
        J Gastrointest Surg. 2008; 12 (discussion: 1716): 1709-1716
        • Eslami M.H.
        • Csikesz N.
        • Schanzer A.
        • Messina L.M.
        Peripheral arterial interventions: trends in market share and outcomes by specialty, 1998-2005.
        J Vasc Surg. 2009; 50: 1071-1078
        • Dua A.
        • Kuy S.
        • Lee C.J.
        • Upchurch Jr., G.R.
        • Desai S.S.
        Epidemiology of aortic aneurysm repair in the United States from 2000 to 2010.
        J Vasc Surg. 2014; 59: 1512-1517