Advertisement

Editor's Choice – Short Term and Long Term Outcomes After Endovascular or Open Repair for Ruptured Infrarenal Abdominal Aortic Aneurysms in the Vascular Quality Initiative

Open ArchivePublished:January 13, 2020DOI:https://doi.org/10.1016/j.ejvs.2019.12.032

      Objective

      Repair of ruptured infrarenal abdominal aortic aneurysms (rAAA) has shifted from open surgical (OAR) to endovascular (EVAR) over the last decade. However, the long term impact of EVAR vs. OAR for rAAA has not been well described.

      Methods

      Prospectively collected registry data (Vascular Quality Initiative [VQI]) were analysed retrospectively to identify patients who underwent EVAR or OAR for rAAA (2004–2018). The primary outcome was death (in hospital and overall post-discharge). Inverse probability weighting (IPW) was used to adjust for treatment selection. Poisson regression assessed the number of one year post-discharge re-interventions.

      Results

      In total, 4257 patients receiving EVAR (n = 2389 [56%]) or OAR (n = 1868 [44%]) for rAAA were identified. Patients were predominantly male (n = 3310 [77.8%]) with a mean ± standard deviation age of 72.7 ± 9.6 years; most (n = 2449 [59.4%]) presented with haemodynamic instability. Use of EVAR for rAAA increased from 7.8% in 2004 to 67.2% in 2018. After IPW, OAR was associated with a higher odds of in hospital mortality (odds ratio [OR] 1.76, 95% confidence interval [CI] 1.54–2.01; p < .001), which was confirmed after multivariable logistic regression (OR 2.08, 95% CI 1.76–2.45; p < .001). Multivariable Cox proportional hazards showed that OAR was also associated with increased overall post-discharge mortality among all patients (hazard ratio 1.36, 95% CI 1.23–1.51; p < .001). Within weighted treatment groups, five year survival was significantly different (55% for EVAR vs. 46% for OAR; p < .001). OAR showed a significantly higher risk of one year post-discharge re-interventions (incidence rate ratio 2.10, 95% CI 1.52–2.89; p < .001).

      Conclusion

      Within the VQI, EVAR for rAAA repair has been increasingly adopted with favourable short term outcomes in terms of morbidity and mortality, as compared with OAR. Unlike elective AAA repair, survival rates between EVAR and OAR do not converge in long term follow up for patients who survived the index hospitalisation.

      Keywords

      Within the Vascular Quality Initiative, endovascular aortic repair (EVAR) for ruptured infrarenal abdominal aortic aneurysm (rAAA) has been adopted increasingly with favourable short term outcomes in terms of morbidity and mortality as compared with open aortic repair (OAR). Unlike elective AAA repair, a convergence of survival rates between EVAR and OAR in long term follow up for patients who survived the index hospitalisation was not observed, suggesting that the early significant benefits of EVAR are sustained over time and an endovascular-first strategy in anatomically feasible candidates with rAAA may be associated with long term benefits.

      INTRODUCTION

      Repair of ruptured infrarenal abdominal aortic aneurysms (rAAA) has shifted from open surgical (OAR) to endovascular aneurysm repair (EVAR) over the last decade, mainly driven by reduced rates of early mortality and morbidity with endovascular treatment.
      • Starnes B.W.
      • Quiroga E.
      • Hutter C.
      • Tran N.T.
      • Hatsukami T.
      • Meissner M.
      • et al.
      Management of ruptured abdominal aortic aneurysm in the endovascular era.
      ,
      • Park B.S.
      • Azefor N.
      • Huang C.
      • Ricotta J.J.
      Trends in treatment of ruptured abdominal aortic aneurysm: impact of endovascular repair and implications for future care.
      A variety of study designs and databases have been used to compare EVAR and OAR for rAAA, and studies of various designs from different databases have reached vastly different conclusions.
      • Robinson W.P.
      Open versus endovascular repair of ruptured abdominal aortic aneurysms: what have we learned after more than 2 decades of ruptured endovascular aneurysm repair?.
      Therefore it remains controversial whether EVAR improves outcomes after rAAA vs. OAR. Indeed, the decision to use EVAR instead of OAR needs to incorporate both early and potential late outcomes, in addition to anatomical restrictions, but the long term outcomes of emergency EVAR for rAAA are not well established. Additionally, the identification of optimal patient subgroups that could benefit from either treatment is poorly defined. Thus, the aim of this study was to contrast the short and long term outcomes after EVAR or OAR for rAAA within a large contemporary national registry.

      MATERIALS AND METHODS

      Data sources and study design

      Prospectively collected data from the Society for Vascular Surgery's Vascular Quality Initiative (VQI) were used to identify patients who had undergone EVAR or OAR for rAAA at participating centres between January 2004, and October 2018. The inclusion criterion for cases was any infrarenal AAA that underwent attempted repair. Patients who were selected to not receive AAA repair (for any reason) or who died prior to incision are not included in the registries. The VQI is a national network of regional quality groups made up of >370 North American academic and community hospitals, with data on > 100 distinct variables for each specific modules (https://www.vqi.org/data-analysis/). Patients having symptomatic or elective AAA repair were excluded.

      Exposures, definitions, and outcomes

      The primary exposure of interest was procedure type, EVAR or OAR. Patients who underwent both OAR and EVAR in the same month and year were assumed to have had the procedures on the same day, assigned as EVAR converted to OAR, and included as EVAR (n = 20). Patients who underwent two OAR or two EVAR procedures in the same month and year were excluded (n = 6), as were patients who underwent EVAR converted to OAR who were recorded only in the open repair registry without a corresponding record in the EVAR registry (n = 38). Patients who underwent repeated aortic procedures during the study period were assigned according to their index operation (n = 5).
      Pre-operative haemodynamic instability was defined as any of the following: lowest blood pressure <90 mmHg, cardiac arrest, or altered mental status. Average annual centre volume of all elective, urgent, and emergency AAA cases was calculated across all years for each institution that contributed at least one record to the VQI registries, excluding the first year of participation and 2018 (as these years were unlikely to include a full year of data). Average annual centre volume was categorised by quartiles.
      The primary outcomes were in hospital mortality and overall post-discharge mortality. Mortality was assessed for all patients by multiple mechanisms, including assessment at clinical follow up in VQI (9–18 months) and matching of patients to the Social Security Death Index and to eligible Medicare claims.
      Secondary outcomes included cardiovascular (myocardial infarction [MI]; new onset congestive heart failure [CHF]; and new onset dysrhythmia) and respiratory complications (pneumonia and re-intubation), post-operative hospital and intensive care unit (ICU) length of stay (LOS), adverse discharge status (any other than home), rates of packed red blood cells (PRBC) transfusions and one year post-discharge aortic re-interventions (captured at one year follow up assessment only).

      Statistical analysis

      Inverse probability weighting (IPW) was used to adjust for treatment selection. This method allows use of the entire data set, giving greater weight to patients with similar characteristics between groups and less weight to those less likely to receive either treatment. To accomplish this, a logistic regression model of undergoing EVAR was developed by including factors with statistical significance or of clinical significance. The IPW was calculated for each patient as the marginal probability of the treatment received divided by the predicted probability of the treatment received derived from the logistic regression model. Patients missing variables used to develop the logistic regression model were excluded (n = 3). Standardised differences were used to assess patient factors between repair types before and after weighting. The standardised difference compares the difference in means in units of the pooled standard deviation (SD) and is not influenced by sample size.
      • Austin P.C.
      Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples.
      Furthermore, it allows for the comparison of the relative balance of variables measured in different units (e.g., years vs. mmHg). A standardised difference < .10 denotes balance in the studied variable.
      Categorical variables were assessed as frequencies. Continuous variables were evaluated as mean ± SD or median (interquartile range) for parametric and non-parametric data, respectively. In the unweighted cohort, differences between groups were calculated using Fisher's exact, Kruskal–Wallis, or chi square tests as appropriate.
      In the weighted cohort, univariable and multivariable models were used to identify factors independently associated with outcomes. Logistic regression was used for in hospital mortality, and Cox proportional hazards for overall post-discharge mortality, along with Kaplan–Meier estimates with log rank test to assess differences. Linear regression assessed hospital and ICU LOS, Poisson regression assessed rates of re-interventions, and PRBC; logistic regression assessed MI, new onset dysrhythmia, new onset CHF, respiratory complications, and unfavourable discharge. Statistical significance was defined as a p value < .05. All statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).

      Ethical approval

      This study was approved by the VQI research advisory council and deemed exempt from approval by the Mayo Clinic Institutional Review Board. All patients’ personal health information was protected, records and outcomes were de-identified, and no testing or procedures were required for this study. Thus, the need for informed consent was waived.

      RESULTS

      Cohort characteristics

      A total of 4257 patients was identified who received EVAR (n = 2389 [56.1%]) or OAR (n = 1868 [43.9%]) for rAAA during the study period (baseline demographics, clinical characteristics, and centre variables in non-weighted treatment groups are reported in Table S1 [Supplementary Material]; procedural details in non-weighted treatment groups are reported in Table S2 [Supplementary Material]). The use of EVAR for rAAA increased from 7.8% in 2004 to 67.2% in 2018 among all rAAA repairs. Prior to weighting, patients who underwent EVAR were older, with smaller AAA, and lower prevalence of haemodynamic instability at presentation. However, no significant differences were found in pre-operative haemoglobin levels. Average centre volume varied across the two approaches; however, similar proportions of EVAR and OAR patients were at the highest quartile volume centres. Patients who had EVAR had a longer median time from symptoms to repair and from admission to repair. Total procedure time was shorter for EVAR. IPW adjustments resulted in similar weighted treatment groups for comparison of outcomes: the two groups were well balanced in terms of baseline demographics and clinical characteristics (Table 1), as well as procedural features (Table 2).
      Table 1Baseline demographics, clinical characteristics, and centre variables before and after inverse probability weighting for endovascular aneurysm repair (EVAR) versus open aneurysm repair (OAR) in patients registered in the Vascular Quality Initiative for treatment of ruptured infrarenal abdominal aortic aneurysms (2004–2018)
      Unweighted statisticsWeighted statistics
      EVAR (n = 2 389)OAR (n = 1 868)Std Diff.EVAR (n = 2 389)OAR (n = 1 853)Std Diff.
      Male sex1 861 (77.9)1 449 (77.6).0081 849 (77.3)1 434 (77.4).002
      Age – y73 (66, 81)72 (66, 79).10573 (66, 80)73 (66, 79).004
      Race
       White2 059 (86.2)1 710 (91.5).1712 116 (88.4)1 639 (88.4)<.001
       Black or African American188 (7.9)72 (3.9).172146 (6.1)111 (6.0).004
       Other/unknown142 (5.9)86 (4.6).060132 (5.5)103 (5.6).004
      BMI – kg/m227.5 (24.0, 31.8)27.5 (24.2, 31.5).03027.7 (24.2, 31.8)27.4 (24.0, 31.2).060
      Hypertension (≥140/90 mmHg or history)1 855 (78.3)1 445 (78.5).0061 858 (78.6%)1 446 (79.2).016
      History of CAD
       None1 837 (77.9)1 415 (77.9).0011 832 (77.8)1 423 (79.0).029
       History MI but no symptoms379 (16.1)291 (16.0).002375 (15.9)281 (15.6).009
       Stable angina88 (3.7)82 (4.5).03994 (4.0)69 (3.8).009
       Unstable angina or MI < 6 mo7 (0.3)7 (0.4).00610 (0.4)4 (0.2).011
       MI < 6 mo35 (1.5)14 (0.8).04633 (1.4)16 (0.9).033
       Unstable angina12 (0.5)8 (0.4).00411 (0.5)8 (0.4).002
      Prior CABG/PCI
       None1 842 (78.2)1 413 (77.6).0121 852 (78.7)1 401 (77.7).023
       PTCA/CABG < 5 y ago153 (6.5)134 (7.4).034159 (6.8)132 (7.3).022
       PTCA/CABG ≥ 5 y ago362 (15.4)273 (15.0).010343 (14.6)269 (14.9).011
      History of CHF
       None2 063 (87.6)1 650 (91.0).1082 080 (89.0)1 619 (89.8).027
       Asymp., hx CHF169 (7.2)98 (5.4).073152 (6.5)111 (6.2).013
       Mild71 (3.0)42 (2.3).04464 (2.7)43 (2.4).022
       Moderate34 (1.4)16 (0.9).03628 (1.2)19 (1.1).007
       Severe17 (0.7)8 (0.4).01814 (0.6)10 (0.6).003
      Pre-operative EF – %
       < 3048 (2.0)33 (1.8).01643 (1.8)37 (2.0).013
       30–50121 (5.1)79 (4.2).040126 (5.3)77 (4.1).054
       > 50279 (11.7)219 (11.7).001257 (10.8)224 (12.1).041
       Not done1 289 (54.0)1 008 (54.0)<.0011 330 (55.6)975 (52.6).060
       Unknown652 (27.3)529 (28.3).023636 (26.6)541 (29.2).058
      Prior CEA/CAS29 (1.4)50 (3.6).14141 (2.1)35 (2.3).014
      Prior bypass48 (2.0)43 (2.4).02046 (2.0)42 (2.3).024
      Prior PVI (PTA/stent)74 (3.1)65 (3.6).02373 (3.1)67 (3.7).035
      Prior major amputation16 (0.7)7 (0.4).01914 (0.6)8 (0.4).008
      Prior aneurysm repair67 (2.9)171 (9.4).271120 (5.2)102 (5.7)0.024
      Diabetes mellitus
       None1 980 (83.9)1 557 (85.1).0341 986 (84.2)1 551 (85.6).038
       Diet91 (3.9)83 (4.5).03491 (3.9)77 (4.2).020
       Non-insulin meds202 (8.6)136 (7.4).041201 (8.5)130 (7.2).051
       Insulin87 (3.7)53 (2.9).04479 (3.3)54 (3.0).022
      Smoking
       Never549 (23.4)312 (17.3).153481 (20.6)372 (20.6)<.001
       Prior839 (35.8)601 (33.2).053792 (33.9)612 (33.9)<.001
       Current957 (40.8)895 (49.5).1751063 (45.5)822 (45.5)<.001
      COPD
       None1 663 (70.7)1 236 (67.7).0651 614 (68.6)1 251 (69.1).012
       Not treated235 (10.0)238 (13.0).095278 (11.8)204 (11.3).016
       On meds314 (13.4)281 (15.4).058348 (14.8)266 (14.7).002
       On home oxygen140 (6.0)71 (3.9).096114 (4.8)88 (4.9).002
      Pre-operative creatinine1.2 (1.0, 1.6)1.2 (1.0, 1.6).0021.2 (1.0, 1.6)1.2 (1.0, 1.6).007
      Pre-operative dialysis
       No2 331 (98.1)1 832 (99.2).0702 340 (98.7)1 816 (98.9).015
       Functioning transplant3 (0.1)6 (0.3).0134 (0.2)4 (0.2).002
       On dialysis41 (1.7)8 (0.4).08327 (1.1)16 (0.9).017
      Pre-operative beta blocker
       No1 264 (54.4)951 (52.6).0361 239 (53.2)961 (53.4).005
       Pre-operative 1–30 d64 (2.8)85 (4.7).10390 (3.9)66 (3.7).012
       Chronic > 30 d842 (36.2)637 (35.2).021829 (35.6)638 (35.5).003
       No, for medical reason32 (1.4)32 (1.8).02543 (1.9)32 (1.8).007
       Operation day only114 (4.9)90 (5.0).003114 (4.9)91 (5.1).008
       Non-compliant9 (0.4)14 (0.8).02512 (0.5)10 (0.5).003
      Pre-operative ACE-I/ARB
       No1 347 (65.5)918 (66.8).0271 261 (65.4)982 (66.0).012
       Yes642 (31.2)408 (29.7).034599 (31.1)463 (31.1).001
       No for medical reason52 (2.5)35 (2.5).00156 (2.9)33 (2.2).045
       Non-compliant14 (0.7)13 (0.9).01713 (0.7)11 (0.7).004
      Pre-operative statin
       No1 272 (54.6)987 (54.7).0011 294 (55.5)958 (53.4).042
       Yes1 000 (42.9)755 (41.8).023974 (41.8)771 (43.0).025
       No for medical reason38 (1.6)40 (2.2).03745 (1.9)42 (2.3).029
       Non-compliant20 (0.9)24 (1.3).03019 (0.8)22 (1.2).026
      Pre-operative antiplatelets980 (42.1)779 (43.1).021953 (40.9)768 (42.8).039
      Pre-operative chronic anticoagulant
       None1 755 (85.1)1 203 (87.4).0681 655 (85.5)1 289 (86.5).028
       Warfarin206 (10.0)102 (7.4).091182 (9.4)130 (8.7).024
       Other67 (3.2)44 (3.2).00358 (3.0)49 (3.3).017
       No, for medical reason30 (1.5)25 (1.8).02336 (1.9)21 (1.4).027
       Non-compliant5 (0.2)2 (0.1).0065 (0.2)2 (0.1).009
      Ambulatory status
       Fully ambulatory522 (84.1)1 266 (90.8).206528 (85.2)1 353 (90.4).160
       Ambulatory with assistance65 (10.5)96 (6.9).12760 (9.7)108 (7.2).086
       Wheelchair21 (3.4)13 (0.9).15718 (2.9)15 (1.0).124
       Bedridden13 (2.1)19 (1.4).04714 (2.3)20 (1.3).058
      Living status
       At home2 298 (96.9)1 806 (97.7).0522 307 (97.2)1 786 (97.6).027
       Nursing home61 (2.6)38 (2.1).03352 (2.2)40 (2.2)<.001
       Homeless13 (0.5)4 (0.2).02115 (0.6)4 (0.2).028
      Lowest pre-intubation BP – mmHg91.0 (70.0, 116.0)84.0 (64.0, 110.0).23090.0 (70.0, 113.0)89.0 (66.0, 112.0).052
      Lowest pre-intubation BP < 90 mmHg1 046 (45.9)993 (55.4).1901 144 (49.9)897 (50.7).016
      Mental status
       Normal1 780 (75.9)1 304 (70.6).1191 754 (74.4)1 337 (73.5).021
       Disoriented332 (14.2)269 (14.6).012329 (13.9)261 (14.4).012
       Unconscious233 (9.9)273 (14.8).148274 (11.6)220 (12.1).015
      Cardiac arrest195 (8.3)253 (13.7).173240 (10.2)192 (10.5).012
      Pre-operative haemodynamic instability1 274 (55.5)1 175 (64.2).1781 374 (59.4)1 076 (59.8).009
      Pre-operative haemoglobin – g/dL11.5 (9.8, 13.2)11.6 (10.0, 13.3).00711.5 (9.8, 13.2)11.6 (10.0, 13.3).014
      Year of procedure
       20044 (0.2)47 (2.5).15029 (1.2)23 (1.2).001
       20054 (0.2)39 (2.1).12336 (1.5)19 (1.0).031
       20069 (0.4)42 (2.2).12021 (0.9)22 (1.2).018
       200710 (0.4)40 (2.1).11026 (1.1)22 (1.2).008
       20089 (0.4)38 (2.0).10624 (1.0)20 (1.1).007
       200921 (0.9)43 (2.3).09136 (1.5)33 (1.8).017
       201047 (2.0)65 (3.5).09364 (2.7)49 (2.6).002
       2011109 (4.6)98 (5.2).032115 (4.8)89 (4.8).001
       2012190 (8.0)144 (7.7).009185 (7.7)146 (7.9).004
       2013269 (11.3)209 (11.2).002258 (10.8)198 (10.7).004
       2014300 (12.6)263 (14.1).045322 (13.4)244 (13.2).008
       2015399 (16.7)252 (13.5).090373 (15.6)286 (15.4).005
       2016377 (15.8)246 (13.2).074341 (14.2)263 (14.2).002
       2017375 (15.7)212 (11.3).127339 (14.2)259 (14.0).006
       2018266 (11.1)130 (7.0).146223 (9.3)181 (9.8).016
      Surgery on weekend
       Monday–Friday1 697 (71.0)1 362 (72.9).0421 704 (71.2)1 361 (73.4).050
       Saturday–Sunday692 (29.0)506 (27.1)689 (28.8)492 (26.6)
      AAA cases/centre/y
       ≤ 1573 (3.1)118 (6.3).15570 (2.9)110 (5.9).145
       > 15–29263 (11.0)235 (12.6).049245 (10.2)254 (13.7).107
       > 29–53634 (26.5)412 (22.1).105598 (25.0)401 (21.6).079
       >531 353 (56.6)1 090 (58.4).0351 421 (59.4)1 072 (57.8).031
       Unable to calculate66 (2.8)13 (0.7).13259 (2.5)16 (0.9).101
      The data are n (%) or median (Q1, Q3) unless stated otherwise. Patients without a known value were excluded from the denominator of that variable. Number of missing values for each variable: body mass index (BMI), n = 308; hypertension, n = 48; history of coronary artery disease (CAD), n = 82; prior coronary artery bypass grafting (CABG)/percutaneous coronary intervention (PCI), n = 80; history (hx) of congestive heart failure (CHF), n = 89; prior carotid endarterectomy (CEA)/carotid artery stenting (CAS), n = 802; prior bypass, n = 75; prior peripheral vascular intervention (PVI) (percutaneous transluminal angioplasty [PTA]/stent), n = 78; prior major amputation, n = 73; prior aneurysm repair, n = 128; diabetes mellitus, n = 68; smoking, n = 104; chronic obstructive pulmonary angioplasty (COPD), n = 79; pre-operative dialysis, n = 36; pre-operative beta blocker, n = 123; pre-operative angiotensin converting enzyme inhibitor (ACE-I)/angiotensin receptor blocker (ARB), n = 828; pre-operative statin, n = 121; pre-operative antiplatelets, n = 121; pre-operative chronic anticoagulant, n = 818; ambulatory status, n = 2 242; living status, n = 37; lowest pre-intubation blood pressure (BP), n = 186; mental status, n = 66; cardiac arrest, n = 58; pre-operative haemodynamic instability, n = 133; pre-operative haemoglobin, n = 287. Std Diff. = standard difference; MI = myocardial infarction; PTCA = percutaneous transluminal coronary angioplasty; Asymp. = asymptomatic; EF = ejection fractionACE-I = angiotensin converting enzyme inhibitor; AAA = abdominal aortic aneurysm.
      Table 2Procedural details before and after inverse probability weighting for endovascular aneurysm repair (EVAR) versus open aneurysm repair (OAR) in patients registered in the Vascular Quality Initiative for treatment of ruptured infrarenal abdominal aortic aneurysms (2004–2018)
      VariablesUnweighted statisticsWeighted statistics
      EVAR (n = 2 389)OAR (n = 1 868)Std Diff.EVAR (n = 2 393)OAR (n = 1 853)Std. Diff.
      Time from symptoms to repair – h8.0 (4.0, 24.0)6.5 (4.0, 18.0).0897.5 (4.0, 21.5)7.0 (4.0, 20.5).053
      Time from admission to repair – hours1.5 (1.0, 3.5)1.3 (0.8, 3.0).0721.5 (0.9, 3.5)1.4 (0.8, 3.0).053
      Pre-operative max. AAA diameter – mm72.0 (60.0, 85.0)78.0 (65.0, 90.0).2274.0 (61.0, 90.0)75.0 (61.0, 90.0).036
      Iliac aneurysm
       No1 693 (76.1)1 412 (78.0).0441 743 (76.7)1 349 (76.6).001
       Unilateral300 (13.5)189 (10.4).094274 (12.0)215 (12.2).005
       Bilateral231 (10.4)210 (11.6).039257 (11.3)196 (11.1).005
      Hypogastric ligated/occluded
       None1 260 (88.8)1 547 (89.4).0201 163 (89.1)1 516 (89.7).017
       Single131 (9.2)115 (6.6).096116 (8.9)114 (6.7).081
       Both28 (2.0)68 (3.9).1226 (2.0)61 (3.6).099
      Anaesthesia
       Local299 (12.6)0 (0).54285 (12.0)0 (0).52
       Regional36 (1.5)0 (0).09741 (1.7)0 (0).11
       General2 038 (85.9)1 866 (100).572 045 (86.3)1 848 (100).56
      Total procedure time – min130.0 (92.0, 180.0)200.0 (150.0, 268.0).75135.0 (95.0, 185.0)202.0 (152.0, 270.0).73
      Data are n (%) or median (Q1, Q3). AAA = abdominal aortic aneurysm; Std Diff. = standard difference.

      Primary outcomes

      Primary and secondary outcomes in non-weighted treatment groups are reported in Table S3 (Supplementary Material). During the study period, overall annual in hospital mortality in the unweighted cohort declined from 29.4% in 2004 to 24.7% in 2018 (p = .034; Table S4 [Supplementary Material]). In weighted treatment groups, OAR was associated with higher odds of in hospital mortality (odds ratio [OR] 1.76, 95% confidence interval [CI] 1.54–2.01; p < .001) (Table 3). Multivariable logistic regression demonstrated that OAR was strongly associated with increased odds of in hospital death compared with EVAR (OR 2.08, 95% CI 1.76–2.45; p < .001) (Table 4). The presence of haemodynamic instability was also significantly predictive of in hospital mortality, with pre-operative cardiac arrest carrying the highest risk (OR 7.07, 95% CI 5.33–9.38; p < .001). Annual AAA volume was associated with reduced risk of in hospital mortality (third quartile volume OR 0.62, 95% CI 0.41–0.93 [p = .021]; highest quartile volume OR 0.69, 95% CI 0.47–1.01 [p = .053]).
      Table 3Univariate comparison of outcomes for endovascular aneurysm repair (EVAR; n = 2 389) versus open aneurysm repair (n = 1 868), in weighted treatment groups, of patients registered in the Vascular Quality Initiative for treatment of ruptured infrarenal abdominal aortic aneurysms
      OutcomesOR (reference: EVAR)95% CIp value
      Primary
       In-hospital mortality1.761.54–2.01<.001
       Overall post-discharge mortality, excluding in-hospital deaths – HR1.321.20–1.45<.001
       Overall post-discharge mortality, limited to patients alive at discharge – HR0.930.77–1.12.44
      Secondary
       Hospital LOS (difference in days) – IRR3.1SE: 0.5<.001
       ICU LOS (difference in days) – IRR3.4SE: 0.3<.001
       MI (troponin or clinical/ECG)1.511.25–1.82<.001
       New onset dysrhythmia1.761.50–2.07<.001
       New onset CHF1.481.15–1.90.002
       Respiratory complications (pneumonia/re-intubation)2.081.81–2.38<.001
       PRBC transfusions – IRR1.601.57–1.64<.001
      Unfavourable discharge (any other than to home)2.251.98–2.56<.001
      Re-interventions – IRR2.101.52–2.89<.001
      Values are reported as odds ratio (OR) unless indicated otherwise. CI = confidence interval; HR = hazard ratio; LOS = length of stay; IRR = incidence rate ratio; ICU = intensive care unit; SE = standard error; MI = myocardial infarction; ECG = electrocardiogram; CHF = congestive heart failure; PRBC = packed red blood cells.
      Table 4Multivariable logistic regression for in-hospital mortality of patients registered in the Vascular Quality Initiative for the treatment of ruptured infrarenal abdominal aortic aneurysms
      ParameterClass levelIn-hospital mortality (OR)95% CIp value
      Treatment group (EVAR vs OAR)EVAR(Ref.)
      OAR2.081.76–2.45<.001
      Average number of cases at centre within 1 y≤15(Ref.)
      >15–290.900.59–1.38.62
      >29–530.620.41–0.93.021
      >530.690.47–1.01.053
      Unable to calculate0.420.20–0.89.024
      Year of surgery2006/20070.900.09–9.44.93
      20080.700.25–2.00.51
      20091.79
      2010(Ref.)0.81–3.93.15
      20110.800.44–1.47.48
      20120.970.55–1.71.93
      20130.710.41–1.23.22
      20140.870.51–1.49.62
      20150.940.56–1.58.81
      20161.130.67–1.92.64
      20171.310.77–2.23.31
      20181.120.65–1.95.68
      SexMale(Ref.)
      Female1.190.98–1.44.08
      Age (per increase of 1 y)1.051.04–1.06<.001
      Prior CHFNone(Ref.)
      Asymp., hx CHF1.160.84–1.62.38
      Mild1.280.76–2.15.36
      Moderate1.480.74–2.93.27
      Severe1.680.59–4.75.33
      Unknown2.600.82–0.32.11
      Pre-operative ejection fraction – %<30(Ref.)
      30–500.510.26–1.00.051
      >500.410.22–0.77.005
      Not done0.480.27–0.88.017
      Unknown0.650.35–0.18.15
      Prior bypassNo(Ref.)
      Yes1.360.80–2.34.26
      0.040.002–0.83.037
      Prior PVI (PTA/stent)No(Ref.)
      Yes2.171.39–3.36<.001
      Unknown22.771.17–443.66.039
      SmokingNever(Ref.)
      Prior0.680.55–0.85<.001
      Current0.760.61–0.95.016
      Unknown2.401.13–5.10.023
      Pre-operative creatinine – per increase of 1 mg/dL1.121.03–1.22.012
      Pre-operative BBNo(Ref.)
      Pre-operatively 1–30 d0.670.37–1.22.19
      Chronic > 30 d1.030.85–1.24.79
      No, for medical reasons1.260.65–2.42.49
      Operation day only0.500.31–0.81.004
      Non-compliant0.580.15–2.17.42
      Unknown12.412.25–68.39.004
      Pre-operative statinNo(Ref.)
      Yes0.830.69–0.99.041
      No for medical reason1.130.61–2.09.69
      Non-compliant1.370.59–3.16.46
      Unknown0.240.04–1.36.11
      Lowest pre-intubation BP < 90 mmHg≥90(Ref.)
      <901.961.65–2.33<.001
      Unknown1.851.19–2.88.007
      Mental statusNormal(Ref.)
      Disoriented1.341.07–1.68.011
      Unconscious1.781.36–2.33<.001
      Unknown1.470.72–2.99.29
      Cardiac arrestNo(Ref.)
      Yes7.075.33–9.38<.001
      Unknown1.700.75–3.85.20
      Pre-operative haemoglobin – per increase of 1 g/dL0.950.92–0.98.004
      OR = odds ratio; CI = confidence interval; EVAR = endovascular aneurysm repair; OAR = open aneurysm repair; Ref. = reference; CHF = congestive heart failure; Asymp. = asymptomatic; hx = history; PVI = peripheral vascular intervention; PTA = percutaneous transluminal angioplasty; BB = beta blocker; BP = blood pressure.
      As shown in Fig. 1A, in weighted treatment groups, overall survival including in hospital deaths was 55% (95% CI 53–58) for EVAR vs. 46% (95% CI 44–49) for OAR (p < .001) five years after surgery. After multivariable adjustment, Cox proportional hazards assessment demonstrated that OAR was associated with increased overall mortality among all patients (hazard ratio [HR] 1.36, 95% CI 1.23–1.51; p < .001) (Table 5). However, when the analysis was limited to patients who were alive at discharge (Fig. 1B), overall mortality in weighted treatment groups was not significantly different at 75% (95% CI 72–78) for EVAR and 75% (95% CI 71–78) for OAR five years after discharge. After multivariable adjustment, Cox proportional hazards assessment demonstrated that survival was similar between treatment groups (HR 0.99, 95% CI 0.79–1.22; p = .90) (Table 6).
      Figure 1
      Figure 1Kaplan–Meier estimates of survival since endovascular aneurysm repair (EVAR) or open aneurysm repair (OAR) for ruptured infrarenal abdominal aortic aneurysm (rAAA) in weighted treatment groups, (A) including in hospital deaths; (B) limited to patients alive at discharge based on Vascular Quality Initiative register data.
      Table 5Multivariable Cox proportional hazards for overall postdischarge mortality (including in-hospital deaths) of patients registered in the Vascular Quality Initiative for treatment of ruptured infrarenal abdominal aortic aneurysm
      ParameterClass levelOverall mortality (HR)95% CIp-value
      Treatment group (EVAR vs OAR)EVAR(Ref.)
      Open AAA1.361.23–1.51<.001
      Average number of cases at centre within 1 y≤15(Ref.)
      >15–290.950.72–1.25.69
      >29–530.690.53–0.89.004
      >530.760.59–0.96.023
      Unable to calculate0.570.34–0.95.032
      Surgery year20062.660.15–47.09.51
      20070.520.10–2.74.44
      20080.940.56–1.59.82
      20091.340.89–2.02.16
      2010(Ref.)
      20110.860.62–1.19.36
      20120.890.65–1.21.44
      20130.720.53–0.98.038
      20140.790.59–1.07.13
      20150.850.63–1.14.28
      20160.970.72–1.31.84
      20171.030.76–1.40.85
      20181.080.78–1.50.65
      SexMale(Ref.)
      Female1.171.03–1.32.012
      Age (per increase of 1 year)1.041.03–1.05<.001
      Prior CHFNone(Ref.)
      Asymp., hx CHF1.180.96–1.45.11
      Mild1.150.83–1.60.41
      Moderate1.350.85–2.12.20
      Severe0.990.48–2.03.98
      Unknown0.920.51–1.66.79
      Pre-operative ejection fraction – %<30(Ref.)
      30–500.690.45–1.06.09
      >500.530.36–0.80.002
      Not done0.640.44–0.93.020
      Unknown0.710.48–1.04.08
      Prior bypassNo(Ref.)
      Yes1.441.01–2.05.043
      Prior PVI (PTA/stent)No(Ref.)
      Yes1.300.98–1.73.07
      Unknown1.930.43–8.57.39
      SmokingNever(Ref.)
      Prior0.780.67–0.89<.001
      Current0.850.74–0.98.027
      Unknown1.050.70–1.55.83
      Preop creatinine – per increase of 1 mg/dL1.030.99–1.09.18
      Pre-operative BBNo(Ref.)
      Pre-operative 1–30 d0.940.66–1.35.75
      Chronic > 30 d1.060.93–1.20.38
      No, for medical reason1.631.09–2.41.016
      Operation day only0.670.49–0.91.011
      Non-compliant1.180.50–2.80.71
      Unknown2.351.05–5.25.037
      Pre-operative statinNo(Ref.)
      Yes0.900.80–1.01.07
      No for medical reason0.930.62–1.40.74
      Non-compliant1.150.63–2.10.65
      Unknown0.880.40–1.95.76
      Lowest pre-intubation BP < 90 mmHg<901.361.22–1.53<.001
      ≥90(Ref.)
      Unknown1.361.03–1.79.031
      Mental statusNormal(Ref.)
      Disoriented1.231.06–1.42.005
      Unconscious1.321.11–1.55.001
      Unknown1.280.79–2.07.32
      Cardiac arrestNo(Ref.)
      Yes3.122.67–3.65<.001
      Unknown1.390.82–2.35.22
      Pre-operative haemoglobin – per increase of 1 g/dL0.960.94–0.98<.001
      HR = hazard ratio; CI = confidence interval; EVAR = endovascular aneurysm repair; OAR = open aneurysm repair; Ref. = reference; CHF = congestive heart failure; Asymp. = asymptomatic; hx = history; PVI = peripheral vascular intervention; PTA = percutaneous transluminal angioplasty; BB = beta blocker; BP = blood pressure.
      Table 6Multivariable Cox proportional hazards for overall postdischarge mortality (limited to patients alive at discharge) of patients registered in the Vascular Quality Initiative for treatment of ruptured infrarenal abdominal aortic aneurysm
      ParameterClass levelOverall mortality after discharge (HR)95% CIp value
      Treatment group (EVAR vs OAR)EVAR(Ref.)
      Open AAA0.990.79–1.22.90
      Average number of cases at centre within 1 year≤15(Ref.)
      >15–29/unable to calculate0.650.33–1.28.21
      >29–530.810.44–1.49.49
      >530.770.43–1.39.39
      Surgery year20070.180.01–3.46.26
      20081.400.65–3.00.39
      20091.800.94–3.44.07
      2010(Ref.)
      20111.150.68–1.93.60
      20120.920.56–1.51.74
      20130.710.42–1.18.18
      20140.710.42–1.18.18
      20150.830.50–1.38.47
      20160.780.46–1.33.36
      20170.370.18–0.76.007
      20180.120.01–1.10.06
      SexMale(Ref.)
      Female1.331.04–1.70.025
      Age – per increase of 1 y1.061.05–1.08<.001
      Prior CHFNone(Ref.)
      Asymp., hx CHF1.310.91–1.90.15
      Mild1.050.53–2.06.89
      Moderate/severe1.390.48–4.01.55
      Unknown2.120.24–18.97.50
      Pre-operative ejection fraction – %<30(Ref.)
      30–501.140.40–3.22.81
      >500.850.31–2.33.76
      Not done0.930.36–2.44.89
      Unknown0.940.36–2.50.91
      Prior bypassNo/unknown(Ref.)
      Yes1.700.75–3.86.20
      Prior PVI (PTA/stent)No/unknown(Ref.)
      Yes0.800.41–1.570.52
      SmokingNever(Ref.)
      Prior0.910.68–1.21.50
      Current1.120.83–1.53.46
      Unknown0.200.03–1.52.12
      Pre-operative creatinine – per increase of 1 mg/dL1.050.94–1.18.38
      Pre-operative BBNo(Ref.)
      Pre-operative 1–30 d1.390.79–2.44.25
      Chronic > 30 d1.160.91–1.49.23
      No, for medical reason3.011.36–6.67.007
      Operation day only1.060.65–1.73.82
      Non-compliant8.381.35–52.21.023
      Unknown0.660.07–6.13.72
      Pre-operative statinNo/unknown(Ref.)
      Yes1.080.86–1.35.53
      No, for medical reason0.240.06–1.04.06
      Non-compliant0.330.03–3.40.35
      Lowest pre-intubation BP < 90 mmHg<900.850.68–1.06.15
      ≥90(Ref.)
      Unknown0.730.37–1.42.35
      Mental statusNormal(Ref.)
      Disoriented1.210.90–1.62.21
      Unconscious1.050.67–1.64.84
      Unknown1.350.31–5.93.69
      Cardiac arrestNo(Ref.)
      Yes1.270.73–2.21.39
      Unknown1.390.33–5.86.65
      Pre-operative haemoglobin – per increase of 1 g/dL0.950.91–1.00.036
      HR = hazard ratio; CI = confidence interval; EVAR = endovascular aneurysm repair; OAR = open aneurysm repair; Ref. = reference; AAA = abdominal aortic aneurysm; CHF = congestive heart failure; Asymp. = asymptomatic; hx = history; PVI = peripheral vascular intervention; PTA = percutaneous transluminal angioplasty; BB = beta blocker; BP = blood pressure.

      Secondary outcomes

      Univariable comparison of outcomes in weighted treatment groups showed that OAR was associated with significantly higher odds of MI (OR 1.51, 95% CI 1.25–1.82; p < .001) and respiratory complications (OR 2.08, 95% CI 1.81–2.38; p < .001), higher rates of blood transfusion (incidence rate ratio [IRR] 1.60, 95% CI 1.57–1.64; p < .001), and longer hospital and ICU LOS (IRR 3.1 [p < .001] and IRR 3.4 [p < .001], respectively). Additionally, OAR showed a significantly higher risk of one year post-discharge re-interventions (IRR 2.10, 95% CI 1.52–2.89 [p < .001]; Table 3).

      DISCUSSION

      AAA rupture remains a dramatically morbid event and vascular surgeons must balance the immediate life saving procedure with the potential for long term efficacy. Current clinical practice guidelines from the European Society for Vascular Surgery strongly recommend EVAR as the first line option for treatment of rAAA when anatomically feasible (Recommendation 74, Class I, Level of Evidence B),
      • Wanhainen A.
      • Verzini F.
      • Van Herzeele I.
      • Allaire E.
      • Bown M.
      • Cohnert T.
      • et al.
      European Society for Vascular Surgery (ESVS) 2019 clinical practice guidelines on the management of abdominal aorto-iliac artery aneurysms.
      mainly owing to the immediate survival advantage. Indeed, the adjusted in hospital mortality from rAAA favoured EVAR with a nearly twofold increased risk of in hospital death with OAR. In the present study, using a large contemporary real world national registry, the use of EVAR has steadily increased over a 15 year period for rAAA repair, which was coupled with a concomitant reduction of the in hospital mortality rate (Fig. 2). Additionally, EVAR was associated with lower rates of peri-operative major adverse events and re-interventions during follow up.
      Figure 2
      Figure 2Graphical representation of in hospital mortality (percentage of in hospital deaths/year of study) vs. endovascular aneurysm repair (EVAR) use for ruptured infrarenal abdominal aortic aneurysm (percentage of EVAR cases/year of study) during the study period (2004–2018).
      These findings are consistent with previous observational studies and systematic reviews,
      • Ten Bosch J.A.
      • Teijinkk J.A.
      • Willigendael E.M.
      • Prins M.H.
      Endovascular aneurysm repair is superior to open surgery for ruptured abdominal aortic aneurysms in EVAR-suitable patients.
      • Gunnarsson K.
      • Wanhainen A.
      • Djavani Gidlund K.
      • Bjorck M.
      • Mani K.
      Endovascular versus open repair as primary strategy for ruptured abdominal aortic aneurysm: a national population-based study.
      • Antoniou G.A.
      • Georgiadis G.S.
      • Antoniou S.A.
      • Pavlidis P.
      • Maras D.
      • Sfyroeras G.S.
      • et al.
      Endovascular repair for ruptured abdominal aortic aneurysm confers an early survival benefit over open repair.
      • Van BeeK S.C.
      • Conijn A.P.
      • Koelemay M.J.
      • Balm R.
      Endovascular aneurysm repair versus open repair for patients with a ruptured abdominal aortic aneurysm: a systematic review and meta-analysis of short-term survival.
      • Qin C.
      • Chen L.
      • Xiao Y.B.
      Emergent endovascular vs. open surgery repair for ruptured abdominal aortic aneurysm: a meta-analysis.
      which have demonstrated significant peri-operative benefits to patients undergoing emergency EVAR for rAAA compared with OAR, especially in those with haemodynamic instability and/or higher pre-operative risk scores.
      • Gupta P.K.
      • Ramanan B.
      • Engelbert T.L.
      • Tefera G.
      • Hoch J.R.
      • Kent K.C.
      A comparison of open surgery versus endovascular repair of unstable ruptured abdominal aortic aneurysms.
      ,
      • Ali M.M.
      • Flahive J.
      • Schanzer A.
      • Simons J.P.
      • Aiello F.A.
      • Doucet D.R.
      • et al.
      In patients stratified by preoperative risk, endovascular repair of ruptured abdominal aortic aneurysms has a lower in-hospital mortality and morbidity than open repair.
      However, an individual patient meta-analysis based on data from three randomised controlled trials (RCTs) concluded that the use of EVAR does not reduce the acute mortality (0–90 days) from rAAA,
      • Sweeting M.J.
      • Balm R.
      • Desgranges P.
      • Ulug P.
      • Powell J.T.
      Ruptured Aneurysm Trialists
      Individual-patient meta-analysis of three randomized controlled trials comparing endovascular versus open repair for ruptured abdominal aortic aneurysm.
      and this was confirmed in a subsequent Cochrane review.
      • Badger S.
      • Forster R.
      • Blair P.H.
      • Ellis P.
      • Kee F.
      • Harkin D.W.
      Endovascular treatment for ruptured abdominal aortic aneurysm.
      Why RCTs and “real world” studies appear to demonstrate conflicting results on the potential benefit of EVAR for rAAA may be related to several factors.
      • Antoniou G.A.
      • Ahmed N.
      • Georgiadis G.S.
      • Torella F.
      Is endovascular repair of ruptured abdominal aortic aneurysm associated with improved in-hospital mortality compared with surgical repair?.
      Firstly, most providers are probably maximising the benefit of EVAR by appropriate patient selection and therefore relegating EVAR to more unstable patients, those presenting to low volume centres without readily available endovascular expertise and devices, or hostile anatomy for EVAR. Alternatively, RCTs may have focused on limited patient subgroups that are not representative of the full range of patients that vascular surgeons treat in clinical practice, thus limiting the external validity of findings to real world practice. Furthermore, OAR for rAAA repair probably results in cumulative adverse peri-operative events, while it is reasonable to assume that the less physiological stress and the faster recovery associated with EVAR are all major factors contributing to better early outcomes.
      In the present study, a net reduction in hospital and ICU LOS, as well as decreased use of blood transfusions, have been demonstrated to be associated with EVAR for rAAA. These findings resemble those of a recent study, which demonstrated a significant reduction in the frequency of prolonged ICU LOS after both elective and non-elective EVAR procedures in the modern endovascular era.
      • Gavali H.
      • Mani K.
      • Tegler G.
      • Cavati R.
      • Covaciu L.
      • Wanhainen A.
      Editor's Choice – prolonged ICU length of stay after AAA repair: analysis of time trends and long-term outcome.
      The clinical importance of transfusions in patients undergoing major vascular surgery is also well documented and still debated, given that many patients have significant underlying cardiovascular morbidity.
      • Chee Y.E.
      • Liu S.E.
      • Irwin M.G.
      Management of bleeding in vascular surgery.
      Whether increased blood loss and more transfusions are causally associated with increased cardiovascular morbidity or are simply a marker of increased cardiovascular risk is unclear.
      • Obi A.T.
      • Park Y.J.
      • Bove P.
      • Cuff R.
      • Kazmers A.
      • Gurm H.S.
      • et al.
      The association of perioperative transfusion with 30-day morbidity and mortality in patients undergoing major vascular surgery.
      ,
      • Osborne Z.
      • Hanson K.
      • Brooke B.S.
      • Schermerhorn M.
      • Henke P.
      • Faizer R.
      • et al.
      Variation in transfusion practices and the association with preioperative adverse events in patients undergoing open abdominal aortic aneurysm repair and lower extremity arterial bypass in the Vascular Quality Initiative.
      Nevertheless, they remain important considerations, and even more so under urgent circumstances and an EVAR-first policy for rAAA (i.e., always EVAR when anatomically suitable) might also result in improved healthcare resource use.
      The likely shortcomings of most available series are the lack of long term follow up, as well as the presence of significant confounders to the analysis. Using robust IPW methodology to account for selection bias, long term post-discharge survival after emergency rAAA repair in the present cohort remained significantly higher during follow up for EVAR compared with OAR. This is in contrast to elective AAA repair where survival after two – four years is similar, with a net loss of the initial survival advantage from EVAR.
      • Schermerhorn M.L.
      • Buck D.B.
      • O'Malley A.J.
      • Curran T.
      • McCallum J.C.
      • Darling J.
      • et al.
      Long-term outcomes of abdominal aortic aneurysm in the Medicare population.
      ,
      • Patel R.
      • Sweeting M.J.
      • Powell J.T.
      • Greenhalgh R.M.
      EVAR Trial Investigators
      Endovascular versus open repair of abdominal aortic aneurysm in 15-years follow-up of the UK endovascular aneurysm repair trial 1 (EVAR trial 1): a randomized controlled trial.
      Nevertheless, the long term findings compare favourably with the three year results of the IMPROVE (Immediate Management of the Patient With Rupture: Open vs. Endovascular Repair) trial where, despite the absence of significant differences between EVAR and OAR in the emergency phase, there were other secondary advantages to EVAR (including shortening the length of hospitalisation and greater likelihood of favourable discharge disposition), which at three years had transformed into a true survival benefit.
      IMPROVE Trial Investigators
      Comparative clinical effectiveness and cost effectiveness of endovascular strategy versus open repair for ruptured abdominal aortic aneurysm: three year results of the IMPROVE randomized trial.
      This confirms previous observations that more patients are able to be discharged home after EVAR and suggests that emergency EVAR may confer significant benefits well beyond the index hospitalisation for rAAA repair.
      • Wallace G.A.
      • Starnes B.W.
      • Hatsukami T.S.
      • Quiroga E.
      • Tang G.L.
      • Kohler T.R.
      • et al.
      Favorable discharge disposition and survival after successful endovascular repair of ruptured abdominal aortic aneurysm.
      Although in the current analysis EVAR did not seem to confer an independent survival benefit in the long run, the early survival gain after EVAR was maintained consistently over time. The cause for this discrepancy is unclear, but some explanations might be offered. Surviving a rAAA may be a physiological stress test, so those surviving are the healthiest and may have the best long term survival.
      • Mani K.
      • Bjorck M.
      • Lundkvist J.
      • Wanhainen A.
      Improved long-term survival after abdominal aortic aneurysm repair.
      By contrast, elective EVAR may have a mix of less healthy patients that survive the procedure but have higher mid term mortality to match those of OAR in the long term. Furthermore, it is acknowledged that the risk of late death after initial rAAA repair is mainly driven by the underlying medical comorbidities and the initial clinical presentation rather than the repair modality itself.
      • Briggs C.S.
      • Sibille J.A.
      • Yammine H.
      • Ballast J.K.
      • Anderson W.
      • Nussbaum T.
      • et al.
      Short-term and mid-term survival of ruptured abdominal aortic aneurysms in the contemporary endovascular era.
      ,
      • von Meijenfledt G.C.I.
      • Ultee K.H.J.
      • Eefting D.
      • Hoeks S.E.
      • ten Raa S.
      • Rouwet E.W.
      • et al.
      Differences in mortality, risk factors, and complications after open and endovascular repair of ruptured abdominal aortic aneurysms.
      However, given the immediate threat to life posed by rAAA, even non-inferiority of long term survival after EVAR would make a valid argument in favour of one treatment over the other.
      A recent report from the IMPROVE trial has reported that although the mid term re-intervention rate after emergency AAA repair (by EVAR or OAR) appears to be twice as high as after elective AAA repair, severe complications were more common after OAR.
      • Powell J.T.
      • Sweeting M.J.
      • Ulug P.
      IMPROVE Trial Investigators
      Editor's Choice – Re-interventions after repair of ruptured abdominal aortic aneurysm. a report from the IMPROVE randomized trial.
      This further suggests that the advantages associated with an endovascular first strategy extend beyond the peri-operative period. In the present study, it was found that EVAR was associated with a lower one year re-intervention rate after the index hospitalisation vs. OAR. This was somewhat unexpected as OAR has traditionally been favoured for the potential for fewer secondary interventions in the long run, and EVAR placed in hostile aortic anatomy would be expected to have increased rates of adverse outcomes and device failure, as the emergency setting may offer placement of a device in marginal anatomy.
      • Baderkhan H.
      • Baston Goncalves F.M.
      • Gomes Oliveira N.
      • Verhagen H.J.M.
      • Wanhainen A.
      • Bjorck M.
      • et al.
      Challenging anatomy predicts mortality and complications after endovascular treatment of ruptured abdominal aortic aneurysm.
      Indeed, the IMPROVE Trial Investigators have conducted an observational study of the treatment received, and not based on the original randomisation, with the aim of identifying morphological parameters that may influence outcomes, and found an inverse relationship between aneurysm neck length and 30 day mortality.
      IMPROVE Trial Investigators
      The effect of aortic morphology on peri-operative mortality of ruptured abdominal aortic aneurysm.
      Yet, the opposite effect was found in the present study, where EVAR was associated with significantly lower rates of post-discharge secondary interventions in early follow up. This may be due to the overall paradigm shift towards endovascular techniques during the last decade and less experience with OAR in the emergency setting.
      • Barshes N.R.
      • McPhee J.
      • Ozaki C.K.
      • Nuyen L.L.
      • Menard M.T.
      • Gravereaux E.
      • et al.
      Increasing complexity in the open surgical repair of abdominal aortic aneurysms.
      However, as short aneurysm necks increase the technical difficulty and complication rates in OAR, and impede conventional EVAR, this may partly explain the benefit of EVAR shown in observational studies but not in RCTs. It is also possible that as VQI tracking for re-interventions is limited to one year with varying follow up by region(s) and centre(s), this may be too early to detect failures needing revisions and this finding may be partly due to a time related bias, which cannot be ascertained. Alternatively, lower re-intervention rates with EVAR may be the result of proper patient selection and increased operator experience. Thus, further work to investigate the causes and types of re-interventions, methods to identify potential “high risk of re-intervention” patients, and more accurate adjustment of results based on baseline aortic anatomy are needed in future studies.
      A recurrent finding of the literature is that the annual hospital volume of AAA procedures is a crucial factor to be considered,
      • Schechter M.A.
      • Pascarella L.
      • Thomas S.
      • McCann R.L.
      • Mureebe L.
      Endovascular and open repair of ruptured infrarenal abdominal aortic aneurysms at a tertiary care center.
      • Trenner M.
      • Kuehnl A.
      • Salvermoser M.
      • Reutersberg B.
      • Geisbuesch S.
      • Schmid v
      • et al.
      Editor's Choice – high annual hospital volume is associated with decreased in-hospital mortality and complication rates following treatment of abdominal aortic aneurysms: secondary data analysis of the nationwide German DRG statistics from 2005 to 2013.
      • Phillips P.
      • Poku E.
      • Essat M.
      • Woods H.B.
      • Goka E.A.
      • Kaltenthaler E.C.
      • et al.
      Procedure volume and the association with short-term mortality following abdominal aortic aneurysm repair in European populations: a systematic review.
      as an inverse relationship with early mortality and morbidity rates seems to exist. In this study, multivariable analysis showed that high annual hospital volume of AAA repairs was associated with a trend towards decreased in hospital mortality. This, combined with the observed benefit from EVAR, would suggest that rAAA would be better centralised to tertiary care centres where protocols, expertise, and equipment for first line EVAR (i.e., EVAR first when feasible) exist but with the possibility of performing OAR if needed. Although it might be argued that relative EVAR and OAR volumes could represent a potential confounder to the analyses, in the present study it was elected to use a pragmatic approach and only include the total AAA volume. In reality, using either EVAR or OAR volumes actually measure the same phenomenon (i.e., a low volume EVAR centre is more likely to be a high volume OAR centre and vice versa), so they are collinear variables. For simplicity, total AAA volume as the metric was selected deliberately. However, broad adoption of a centralised EVAR first approach for the treatment of rAAA requires massive systemic changes, and the question of which patients with rAAA may wait to be transported to a centre that can offer both open and endovascular repair remains unclear. Indeed, despite the increasing evidence that regionalisation of emergency vascular surgery for patients with rAAA improves outcomes,
      • Warner C.J.
      • Roddy S.P.
      • Chang B.B.
      • Kreienberg P.P.
      • Sternbach Y.
      • Taggert J.B.
      • et al.
      Regionalization of emergent vascular surgery for patients with ruptured AAA improves outcomes.
      many rAAA repairs are still being performed in low volume centres and/or those offering a primary OAR strategy.
      • Budtz-Lilly J.
      • Bjorck M.
      • Venermo M.
      • Debus S.
      • Behrendt C.A.
      • Altreuther M.
      • et al.
      Editor's Choice – the impact of centralization and endovascular aneurysm repair on treatment of ruptured abdominal aortic aneurysms based on international registries.
      Therefore, future work is needed to improve regionalisation and standardisation of acute aortic care.

      Limitations

      This study has limitations. Firstly, it was a retrospective analysis of real world data and cannot replace a well designed RCT. However, to the extent possible, comparable groups were created using IPW techniques to adjust for selection bias and multivariable analysis to account for known residual confounders. Owing to limitations with Social Security Death Index data and patient follow up, mortality assessment may be under reported; however, it is not expected that this should differ by group and therefore is unlikely to alter the findings. It is acknowledged that there are no data from patients who died before surgery, and it is also not known if all patients were eligible for EVAR (on/off instructions for use) as certain anatomical data needed to determine this are missing in the open AAA data set. Indeed, neck diameter, length, and angulation are not collected in the open AAA forms for VQI; they are only in the EVAR forms. Also, tracking this information within EVAR forms was initiated around 2014, so it is not uniformly available. Thus, it was decided to leave this information out of the analysis. The data do not capture how parallel grafting or other advanced endovascular techniques may augment treatment of juxtarenal/pararenal rAAA. However, all patients in both groups, by the VQI inclusion definition, have an infrarenal AAA. Thus, the clamp position in rAAA is not a surrogate for aneurysm extent in this setting, but probably a means of the most effective immediate aortic clamping and haemorrhage control. Lastly, indications for re-interventions in VQI are broad for both EVAR (sac growth, endoleak, migration, occlusion, stenosis, rupture, graft infection) and OAR (incision, graft, intestine, leg ischaemia). However, they may not be able to capture the entire spectrum of possible complications requiring secondary treatment.

      CONCLUSION

      Within the VQI, EVAR for rAAA repair has been increasingly adopted with favourable short term outcomes in terms of morbidity and mortality vs. OAR. Unlike elective AAA repair, survival rates between EVAR and OAR were not observed to converge in long term follow up for patients who survived the index hospitalisation. This suggests that the early significant benefits of EVAR are sustained over time and promotes an endovascular first strategy in anatomically feasible candidates with rAAA.

      Acknowledgements

      Presented as part of the Poster Competition at the 2019 Society for Vascular Surgery Vascular Annual Meeting, National Harbor, MD, USA.

      Appendix A. Supplementary data

      The following are the Supplementary data to this article:

      Conflict of Interest

      None.

      FUNDING

      None.

      References

        • Starnes B.W.
        • Quiroga E.
        • Hutter C.
        • Tran N.T.
        • Hatsukami T.
        • Meissner M.
        • et al.
        Management of ruptured abdominal aortic aneurysm in the endovascular era.
        J Vasc Surg. 2010; 51: 9-18
        • Park B.S.
        • Azefor N.
        • Huang C.
        • Ricotta J.J.
        Trends in treatment of ruptured abdominal aortic aneurysm: impact of endovascular repair and implications for future care.
        J Am Coll Surg. 2013; 216: 745-755
        • Robinson W.P.
        Open versus endovascular repair of ruptured abdominal aortic aneurysms: what have we learned after more than 2 decades of ruptured endovascular aneurysm repair?.
        Surgery. 2017; 162: 1207-1218
        • Austin P.C.
        Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples.
        Stat Med. 2009; 28: 3083-3107
        • Wanhainen A.
        • Verzini F.
        • Van Herzeele I.
        • Allaire E.
        • Bown M.
        • Cohnert T.
        • et al.
        European Society for Vascular Surgery (ESVS) 2019 clinical practice guidelines on the management of abdominal aorto-iliac artery aneurysms.
        Eur J Vasc Endovasc Surg. 2019; 57: 8-93
        • Ten Bosch J.A.
        • Teijinkk J.A.
        • Willigendael E.M.
        • Prins M.H.
        Endovascular aneurysm repair is superior to open surgery for ruptured abdominal aortic aneurysms in EVAR-suitable patients.
        J Vasc Surg. 2010; 52: 213-218
        • Gunnarsson K.
        • Wanhainen A.
        • Djavani Gidlund K.
        • Bjorck M.
        • Mani K.
        Endovascular versus open repair as primary strategy for ruptured abdominal aortic aneurysm: a national population-based study.
        Eur J Vasc Endovasc Surg. 2016; 51: 22-28
        • Antoniou G.A.
        • Georgiadis G.S.
        • Antoniou S.A.
        • Pavlidis P.
        • Maras D.
        • Sfyroeras G.S.
        • et al.
        Endovascular repair for ruptured abdominal aortic aneurysm confers an early survival benefit over open repair.
        J Vasc Surg. 2013; 58: 1091-1095
        • Van BeeK S.C.
        • Conijn A.P.
        • Koelemay M.J.
        • Balm R.
        Endovascular aneurysm repair versus open repair for patients with a ruptured abdominal aortic aneurysm: a systematic review and meta-analysis of short-term survival.
        Eur J Vasc Endovasc Surg. 2014; 47: 593-602
        • Qin C.
        • Chen L.
        • Xiao Y.B.
        Emergent endovascular vs. open surgery repair for ruptured abdominal aortic aneurysm: a meta-analysis.
        PLoS One. 2014; 9e87465
        • Gupta P.K.
        • Ramanan B.
        • Engelbert T.L.
        • Tefera G.
        • Hoch J.R.
        • Kent K.C.
        A comparison of open surgery versus endovascular repair of unstable ruptured abdominal aortic aneurysms.
        J Vasc Surg. 2014; 60: 1439-1445
        • Ali M.M.
        • Flahive J.
        • Schanzer A.
        • Simons J.P.
        • Aiello F.A.
        • Doucet D.R.
        • et al.
        In patients stratified by preoperative risk, endovascular repair of ruptured abdominal aortic aneurysms has a lower in-hospital mortality and morbidity than open repair.
        J Vasc Surg. 2015; 61: 1399-1407
        • Sweeting M.J.
        • Balm R.
        • Desgranges P.
        • Ulug P.
        • Powell J.T.
        • Ruptured Aneurysm Trialists
        Individual-patient meta-analysis of three randomized controlled trials comparing endovascular versus open repair for ruptured abdominal aortic aneurysm.
        Br J Surg. 2015; 102: 1229-1239
        • Badger S.
        • Forster R.
        • Blair P.H.
        • Ellis P.
        • Kee F.
        • Harkin D.W.
        Endovascular treatment for ruptured abdominal aortic aneurysm.
        Cochrane Database Syst Rev. 2017; 5CD005261
        • Antoniou G.A.
        • Ahmed N.
        • Georgiadis G.S.
        • Torella F.
        Is endovascular repair of ruptured abdominal aortic aneurysm associated with improved in-hospital mortality compared with surgical repair?.
        Interact Cardiovasc Thorac Surg. 2015; 20: 135-139
        • Gavali H.
        • Mani K.
        • Tegler G.
        • Cavati R.
        • Covaciu L.
        • Wanhainen A.
        Editor's Choice – prolonged ICU length of stay after AAA repair: analysis of time trends and long-term outcome.
        Eur J Vasc Endovasc Surg. 2017; 54: 157-163
        • Chee Y.E.
        • Liu S.E.
        • Irwin M.G.
        Management of bleeding in vascular surgery.
        Br J Anaesth. 2016; 117 (ii85–94)
        • Obi A.T.
        • Park Y.J.
        • Bove P.
        • Cuff R.
        • Kazmers A.
        • Gurm H.S.
        • et al.
        The association of perioperative transfusion with 30-day morbidity and mortality in patients undergoing major vascular surgery.
        J Vasc Surg. 2015; 61: 1000-1009
        • Osborne Z.
        • Hanson K.
        • Brooke B.S.
        • Schermerhorn M.
        • Henke P.
        • Faizer R.
        • et al.
        Variation in transfusion practices and the association with preioperative adverse events in patients undergoing open abdominal aortic aneurysm repair and lower extremity arterial bypass in the Vascular Quality Initiative.
        Ann Vasc Surg. 2018; 46: 1-16
        • Schermerhorn M.L.
        • Buck D.B.
        • O'Malley A.J.
        • Curran T.
        • McCallum J.C.
        • Darling J.
        • et al.
        Long-term outcomes of abdominal aortic aneurysm in the Medicare population.
        N Engl J Med. 2015; 373: 328-338
        • Patel R.
        • Sweeting M.J.
        • Powell J.T.
        • Greenhalgh R.M.
        • EVAR Trial Investigators
        Endovascular versus open repair of abdominal aortic aneurysm in 15-years follow-up of the UK endovascular aneurysm repair trial 1 (EVAR trial 1): a randomized controlled trial.
        Lancet. 2016; 388: 2366-2374
        • IMPROVE Trial Investigators
        Comparative clinical effectiveness and cost effectiveness of endovascular strategy versus open repair for ruptured abdominal aortic aneurysm: three year results of the IMPROVE randomized trial.
        BMJ. 2017; 359j4859
        • Wallace G.A.
        • Starnes B.W.
        • Hatsukami T.S.
        • Quiroga E.
        • Tang G.L.
        • Kohler T.R.
        • et al.
        Favorable discharge disposition and survival after successful endovascular repair of ruptured abdominal aortic aneurysm.
        J Vasc Surg. 2013; 57: 1495-1502
        • Mani K.
        • Bjorck M.
        • Lundkvist J.
        • Wanhainen A.
        Improved long-term survival after abdominal aortic aneurysm repair.
        Circulation. 2009; 120: 201-211
        • Briggs C.S.
        • Sibille J.A.
        • Yammine H.
        • Ballast J.K.
        • Anderson W.
        • Nussbaum T.
        • et al.
        Short-term and mid-term survival of ruptured abdominal aortic aneurysms in the contemporary endovascular era.
        J Vasc Surg. 2018; 68: 408-415
        • von Meijenfledt G.C.I.
        • Ultee K.H.J.
        • Eefting D.
        • Hoeks S.E.
        • ten Raa S.
        • Rouwet E.W.
        • et al.
        Differences in mortality, risk factors, and complications after open and endovascular repair of ruptured abdominal aortic aneurysms.
        Eur J Vasc Endovsc Surg. 2014; 47: 479-486
        • Powell J.T.
        • Sweeting M.J.
        • Ulug P.
        • IMPROVE Trial Investigators
        Editor's Choice – Re-interventions after repair of ruptured abdominal aortic aneurysm. a report from the IMPROVE randomized trial.
        Eur J Vasc Endovasc Surg. 2018; 55: 625-632
        • Baderkhan H.
        • Baston Goncalves F.M.
        • Gomes Oliveira N.
        • Verhagen H.J.M.
        • Wanhainen A.
        • Bjorck M.
        • et al.
        Challenging anatomy predicts mortality and complications after endovascular treatment of ruptured abdominal aortic aneurysm.
        J Endovasc Ther. 2016; 23: 919-927
        • IMPROVE Trial Investigators
        The effect of aortic morphology on peri-operative mortality of ruptured abdominal aortic aneurysm.
        Eur Heart J. 2015; 36: 1328-1334
        • Barshes N.R.
        • McPhee J.
        • Ozaki C.K.
        • Nuyen L.L.
        • Menard M.T.
        • Gravereaux E.
        • et al.
        Increasing complexity in the open surgical repair of abdominal aortic aneurysms.
        Ann Vasc Surg. 2012; 26: 10-17
        • Schechter M.A.
        • Pascarella L.
        • Thomas S.
        • McCann R.L.
        • Mureebe L.
        Endovascular and open repair of ruptured infrarenal abdominal aortic aneurysms at a tertiary care center.
        Ann Vasc Surg. 2017; 41: 83-88
        • Trenner M.
        • Kuehnl A.
        • Salvermoser M.
        • Reutersberg B.
        • Geisbuesch S.
        • Schmid v
        • et al.
        Editor's Choice – high annual hospital volume is associated with decreased in-hospital mortality and complication rates following treatment of abdominal aortic aneurysms: secondary data analysis of the nationwide German DRG statistics from 2005 to 2013.
        Eur J Vasc Endovasc Surg. 2018; 55: 185-194
        • Phillips P.
        • Poku E.
        • Essat M.
        • Woods H.B.
        • Goka E.A.
        • Kaltenthaler E.C.
        • et al.
        Procedure volume and the association with short-term mortality following abdominal aortic aneurysm repair in European populations: a systematic review.
        Eur J Vasc Endovasc Surg. 2017; 53: 77-88
        • Warner C.J.
        • Roddy S.P.
        • Chang B.B.
        • Kreienberg P.P.
        • Sternbach Y.
        • Taggert J.B.
        • et al.
        Regionalization of emergent vascular surgery for patients with ruptured AAA improves outcomes.
        Ann Surg. 2016; 264: 538-543
        • Budtz-Lilly J.
        • Bjorck M.
        • Venermo M.
        • Debus S.
        • Behrendt C.A.
        • Altreuther M.
        • et al.
        Editor's Choice – the impact of centralization and endovascular aneurysm repair on treatment of ruptured abdominal aortic aneurysms based on international registries.
        Eur J Vasc Endovasc Surg. 2018; 56: 181-188

      Comments

      Commenting Guidelines

      To submit a comment for a journal article, please use the space above and note the following:

      • We will review submitted comments as soon as possible, striving for within two business days.
      • This forum is intended for constructive dialogue. Comments that are commercial or promotional in nature, pertain to specific medical cases, are not relevant to the article for which they have been submitted, or are otherwise inappropriate will not be posted.
      • We require that commenters identify themselves with names and affiliations.
      • Comments must be in compliance with our Terms & Conditions.
      • Comments are not peer-reviewed.