Publications

OBJECTIVES: Physician-modified endografts (PMEGs) have expanded the scope of endovascular abdominal aortic repair beyond the infrarenal aorta. Patients with prohibitively high surgical risk and visceral segment disease are often candidates for this intervention, which mitigates much of the morbidity and mortality associated with conventional open repair. Here we present the institutional PMEG experience of a high-volume aortic center.

METHODS: We studied all PMEGs performed at our institution from 2012 to 2023. We included cases that were submitted to the US Food sand Drug Administration in support of an investigational device exemption (IDE) trial, as well as those in the subsequently approved IDE trial. Over this 11-year period, we assessed the changes in operative characteristics and perioperative outcomes over time. Additionally, we compared the outcomes from PMEG cases to those of Zenith fenestrated (ZFEN) grafts (done by the surgeon with the PMEG IDE), an alternative device used for aneurysms involving the lower visceral segment. Here we assessed operative characteristics, perioperative outcomes, and 5-year survival and reintervention rates.

RESULTS: When assessing the change over time for PMEG operative characteristics, we found a trend toward decreased fluoroscopy time and decreased proportions of completion type I and type III endoleaks (all P < .05). Perioperative outcomes have remained stable over this period, with an overall perioperative mortality rate of 4.9% (noting that this registry also includes cases that were urgent and emergent). Despite the increased complexity of PMEGs relative to ZFENs, we found comparable perioperative outcomes with regard to mortality (4.9% vs 4.3%; P = .86), permanent spinal cord ischemia (1.1% vs 0%; P = .38), postoperative myocardial infarction (4.3% vs 2.9%; P = .60), postoperative respiratory failure (7.1% vs 4.3%; P = .43), and new dialysis use (2.2% vs 4.3%; P = .35). Additionally, 5-year survival (PMEG 54% vs ZFEN 65%; P = .15) and freedom from reintervention (63% vs 74%; P = .07) were similar between these cohorts.

CONCLUSIONS: Throughout our >10-year experience with PMEGs, we have noted improvements in operative outcomes, which can likely be attributed to technological advances and increased physician experience. Additionally, we have found that PMEGs perform well when compared with ZFENs, despite being a more complicated repair that is able to treat a larger segment of the aorta. PMEGs are crucial for the comprehensive care of vascular patients with complex aortic disease. As further operative advancements are made, we only expect the use of this intervention to increase.

OBJECTIVE: The aim of this study was to compare perioperative and 5-year outcomes following endovascular (FEVAR) and open repair (OAR) of complex abdominal aortic aneurysms (cAAAs) in males and females separately, given the known sex-related differences in perioperative outcomes.

METHODS: We studied all elective cAAA repairs between 2014 and 2019 in the Vascular Implant Surveillance and Interventional Outcomes Network (VISION) registry. We stratified patients based on sex. We calculated propensity scores for assignment to either OAR or FEVAR. Covariates including age, race, diameter, baseline comorbidities, proximal extent of repair, annual center volumes, and annual surgeon volumes were introduced into the model for estimating propensity scores. Within matched cohorts, perioperative outcomes and 5-year outcomes (mortality, reinterventions, and ruptures) were evaluated using multivariable logistic and Cox regression models.

RESULTS: We identified 2825 patients, of whom 29% were female. Within both the sexes, OAR was more commonly performed (OAR vs FEVAR: males: 53% vs 47%; females: 63% vs 37%). After matching, among males (n = 1326), FEVAR was associated with lower perioperative mortality (FEVAR vs OAR: 2.3% vs 5.1%; P < .001). However, FEVAR was associated with comparable 5-year mortality (38% vs 28%; hazard ratio [HR], 1.2; 95% confidence interval [CI], 0.92-1.4; P = .22) and a higher hazard of 5-year reintervention (19% vs 3.7%; adjusted HR, 4.5; 95% CI, 2.6-7.6; P < .001). Among females (n = 456), FEVAR and OAR showed similar perioperative mortality (8.3% vs 7.0%; P = .73). At 5 years, FEVAR was associated with higher hazards of mortality (43% vs 32%; adjusted HR, 1.5; 95% CI, 1.03-2.2; P = .034) and reintervention (20% vs 3.0%; adjusted HR, 4.8; 95% CI, 2.1-11; P < .001) compared with OAR.

CONCLUSIONS: Among males, FEVAR was associated with favorable perioperative outcomes compared with OAR, although these advantages attenuate over time. However, among females, FEVAR was associated with similar perioperative outcomes, eventually leading to higher reinterventions and possibly higher mortality within 5 years. Future efforts should focus on determining the factors associated with these sex disparities to improve outcomes following FEVAR in females. Based on current evidence, females undergoing elective cAAA repair should be selected with due caution, especially for endovascular repair.

OBJECTIVE: Several studies comparing the transperitoneal (TP) and retroperitoneal (RP) approach for abdominal aortic aneurysm (AAA) repair suggest that the RP approach may result in lower rates of perioperative mortality and morbidity. However, data comparing these approaches for open conversion are lacking. This study aims to evaluate the association between the type of approach and outcomes following open conversion after endovascular aneurysm repair (EVAR).

METHODS: We included all patients who underwent open conversion after EVAR between 2010 and 2022 in the Vascular Quality Initiative. Patients presenting with rupture were excluded. The primary outcome was perioperative mortality. The secondary outcomes included perioperative complications and 5-year mortality. Inverse probability weighting was used to adjust for factors with statistical or clinical significance. Logistic regression was used to assess perioperative mortality and complications in the weighted cohort. The 5-year mortality was evaluated using Kaplan-Meier and Cox regression.

RESULTS: We identified 660 patients (39% RP) who underwent open conversion after EVAR. Compared with TP, RP patients were older (75 years [interquartile range, 70-79 years] vs 73.5 years [interquartile range, 68-79 years]; P < .001), and more frequently had prior myocardial infarction (33% vs 22%; P = .002). Compared with the TP approach, the RP approach was used less frequently in cases of associated iliac aneurysm (19% vs 27%; P = .026), but more frequently with associated renal bypass (7.8% vs 1.7%; P < .001) and by high-volume physicians (highest quintile, >7 AAA annually: 41% vs 17%; P < .001) and in high-volume centers (highest quintile, >35 AAA annually: 36% vs 20%; P < .001). RP patients, compared with TP patients, were less likely to have external iliac or femoral distal anastomosis (8.2% vs 21%; P < .001), and an infrarenal clamp (25% vs 36%; P < .001). Unadjusted perioperative mortality was not significantly different between approaches (RP vs TP: 3.8% vs 7.5%; P = .077). After risk adjustment, RP patients had similar odds of perioperative mortality (adjusted odds ratio [aOR], 0.49; 95% confidence interval [CI], 0.22-1.10; P = .082), and lower odds of intestinal ischemia (aOR, 0.26; 95% CI, 0.08-0.86; P = .028) and in-hospital reintervention (aOR, 0.43; 95% CI, 0.22-0.85; P = .015). No significant differences were found in the other perioperative complications or 5-year mortality (aHR, 0.79; 95% CI, 0.47-1.32; P = .37).

CONCLUSIONS: Our findings suggest that the RP approach may be associated with a lower adjusted odds of perioperative complications compared with the TP approach. The RP approach should be considered for open conversion after EVAR when feasible.

BACKGROUND: Drug-coated balloons and drug-eluting stents (DCB/DES) have shown promise in improving outcomes for patients with peripheral artery disease (PAD); however, more real-world analyses are needed to better understand the role of this technology within current practice. As such, we compared our institution's experience with DCB/DES vs percutaneous transluminal angioplasty with or without stenting (PTA/S) for the treatment of PAD.

METHODS: All patients undergoing an infrainguinal endovascular intervention for PAD at our institution between 2016 and 2022 were reviewed retrospectively. Patients undergoing isolated supra-inguinal or tibial interventions were excluded. Outcomes included primary patency, freedom from major adverse limb events (MALE), limb salvage, and amputation-free survival. To account for baseline differences, one-to-one propensity score matching was performed between DCB/DES and PTA/S groups. Outcomes were further evaluated using χ2, Kaplan-Meier analyses, and Cox regression analyses.

RESULTS: Between 2016 and 2022, 800 patients underwent an endovascular infrainguinal intervention for PAD: 224 DCB/DES and 576 PTA/S. Before matching, DCB/DES patients were younger (69 vs 72 years), more often male (72% vs 65%), non-White (58%, vs 34%), had a history of smoking (77% vs 64%), were more likely to have had a prior ipsilateral intervention (45% vs 10%), and more often presented with claudication (44% vs 18%) (all P < .05). After matching, 211 patients were included in each group, where the only remaining difference between DCB/DES and PTA/S was prior ipsilateral intervention (40% vs 15%; P < .001). After DCB/DES, Kaplan-Meier analyses demonstrated higher rates of primary patency (65% vs 54%; P < .01) and higher freedom from MALE (3-year rates: 84% vs 75%; P = .04), correlating with a 38% lower event risk in both outcomes (hazard ratio [HR], 0.62 [95% confidence interval, 0.44-0.89] and hazard ratio, 0.62 [0.39-0.99], respectively). No differences were noted in rates of limb salvage (3-year rates: 94% vs 90%; P = .63) or amputation-free survival (3-year rates: 78% vs 71%; P = .13). When stratifying by indication, DCB/DES demonstrated higher rates of freedom from MALE among patients with chronic limb-threatening ischemia (3-year rates: 85% vs 66%; P = .02).

CONCLUSIONS: Among a matched cohort of patients with PAD undergoing endovascular intervention, DCB/DES, as compared with PTA/S, demonstrated higher rates of primary patency and freedom from MALE, the former treatment effect remaining notable among patients with claudication and the latter among patients with chronic limb-threatening ischemia. These data demonstrate the importance of further analyses on this evolving technology.

OBJECTIVE: Outcomes following redo carotid endarterectomy (rCEA) have been shown to be worse than those after primary CEA (pCEA). Additional research has shown that outcomes are better with transcarotid artery revascularization (TCAR) for restenosis after CEA compared with rCEA and transfemoral carotid artery stenting; however, not all patients are eligible for TCAR or transfemoral carotid artery stenting. Given the increasing utilization of endovascular techniques, this study aims to evaluate changes in outcomes of rCEA vs pCEA before and after the approval of TCAR by the United States Food and Drug Administration in 2015.

METHODS: All patients between 2003 and 2023 who underwent CEA in the Vascular Quality Initiative were included and categorized as pCEA or rCEA. Cochrane-Armitage trend testing was used to examine trends in proportion of rCEA compared with pCEA, and the Mann-Kendall trend test was used for perioperative outcomes following rCEA overtime. Multivariable logistic regression was used to compare in-hospital stroke/death, stroke, death, and stroke/death/myocardial infarction following rCEA vs pCEA after stratifying patients into two cohorts: 2003 to 2015 and 2016 to 2023 (before and after introduction of TCAR). Analysis was also performed based on preoperative symptoms.

RESULTS: Of 198,150 patients undergoing CEA, 98.4% were pCEA and 1.6% were rCEA. During the study period, the proportion of rCEA in the Vascular Quality Initiative decreased from 2.3% to 1.0% as endovascular methods became more available (P < .001). Trend testing of individual outcomes showed an increase in the stroke/death rate following rCEA over time (P = .019) despite an improvement in the death rate (P = .009). From 2003 to 2015, patients undergoing rCEA had higher odds of stroke/death compared with pCEA (2.4% vs 1.2%; adjusted odds ratio [aOR], 1.81; 95% confidence interval [CI], 1.14-2.73; P = .007). Higher stroke/death rates after rCEA persisted only in asymptomatic patients (2.3% vs 1.1%; aOR, 2.03; 95% CI, 1.19-3.25; P = .006); however, there was no difference in symptomatic patients (3.0% vs 2.0%; aOR, 1.37; 95% CI, 0.51;3.01; P = .50). In the late period, rCEA had higher odds of stroke/death compared with pCEA (3.1% vs 1.3%; aOR, 2.45; 95% CI, 1.85-3.18; P < .001), and the association was seen in asymptomatic patients (1.9% vs 1.0%; aOR, 1.95; 95% CI, 1.29-2.82; P < .001) and symptomatic patients (6.3% vs 2.0%; aOR, 3.23; 95% CI, 2.17-4.64; P < .001).

CONCLUSIONS: The proportion of rCEAs done yearly in the United States has been decreasing as endovascular options became available. As the rate of rCEA has decreased, outcomes have been worsening, with an increasing stroke/death rate seen over time, driven primarily by worse outcomes in symptomatic patients. Stroke/death rates for asymptomatic patients fall within Society for Vascular Surgery guidelines, and so the choice between rCEA, CAS, or medical management should be made after shared decision-making between a patient and their surgeon. However, with an in-hospital stroke death rate of over 6% symptomatic patients should be selected very carefully, as some are less likely to benefit from rCEA.

OBJECTIVE: Because aneurysmal disease is progressive, proximal disease progression and para-anastomotic aneurysms are complications experienced after open infrarenal abdominal aortic aneurysm (AAA) repair. As such, fenestrated or branched endovascular repair (F/BEVAR) may be indicated in these patients. Data describing fenestrated endovascular aneurysm repair after prior open surgical repair (OSR) are limited to institutional databases. The aim of our study is to describe the safety and efficacy of fenestrated/branched endovascular aneurysm repair (F/BEVAR) in patients with prior OSR compared with primary F/BEVAR using the Vascular Quality Initiative.

METHODS: Using the VQI complex endovascular AAA module from 2014 to 2022, we identified all single-staged F/BEVAR repair in patients having prior OSR or no prior aortic surgery (primary F/BEVAR). The primary outcomes were perioperative mortality and completion endoleaks. Secondary outcomes were 5-year survival and 1-year sac dynamics. Between the two cohorts, differences in the primary and secondary outcomes were evaluated using Wilcoxon rank-sum tests for continuous variables and χ2 analysis for categorical variables. Kaplan-Meier methods and Cox regression were used to examine 5-year mortality.

RESULTS: We identified 3331 primary F/BEVAR patients and 102 prior OSR patients. Patients with prior OSR were more likely to have peripheral arterial disease (22.0% vs 7.4%), prior smoking (67% vs 56%), and undergo F/BEVAR with medium-/high-volume physicians (74% vs 62%), but less likely to be female (8.8% vs 23.0%) (all P < .05). Patients with prior OSR were also more likely to have a more proximal aneurysm extent (median zone 7 [interquartile range (IQR), zones 6-8] vs zone 8 [IQR, zones 7-8]), larger AAA diameters (62 mm [IQR, 56-66 mm] vs 58 mm [IQR, 55-63 mm]), receive a physician-modified endograft (PMEG) vs commercial custom-made device (36% vs 20% physician-modified endograft), have longer surgery times (240 minutes [IQR, 186-308 minutes] vs 206 minutes [IQR, 155-272 minutes]), and have a higher rate of celiac artery (51% vs 26%) and superior mesenteric artery (86% v 73%) artery involvement (all P < .05). Patients with prior OSR had lower rates of completion endoleaks (25% vs 36%), driven by lower rates of type II leaks (11% vs 20%) despite higher rates of indeterminate leaks (11% vs 5.1%) (all P < .01). There was, however, no difference in perioperative mortality (2% vs 2.9%; P = .78). They had similar 1-year sac dynamics (48% vs 50% regression; 12% vs 8% expansion; P > .5) and 5-year mortality (23% vs 18%; hazard ratio [HR], 1.44 [IQR, 0.89-2.31]; P = .13).

CONCLUSIONS: Based on VQI data, F/BEVAR after prior OSR seems to be well-tolerated and safe. Prior OSR patients also had lower rates of completion type II endoleaks and similar sac dynamics and 5-year mortality compared with primary F/BEVAR patients.

OBJECTIVE: BEST-CLI established the superiority of single-segment great saphenous vein (ssGSV) conduits for revascularization in patients with chronic limb-threatening ischemia (CLTI); however, the generalizability of these data is unknown. Thus, we aimed to validate the long-term results of open surgical bypass (BPG) vs percutaneous transluminal angioplasty with or without stenting (PTA/S) using the BEST-CLI inclusion and randomization criteria.

METHODS: All patients undergoing a first-time lower extremity revascularization for CLTI at our institution from 2005 to 2022 were retrospectively reviewed. To approximate BEST-CLI, one-to-one propensity score matching was used. Cohort 1 included BPG with ssGSV vs PTA/S; Cohort 2 included BPG without ssGSV vs PTA/S. Primary outcomes included wound healing, major amputation, major reintervention, major amputation/death (amputation/death), and major adverse limb events (MALE) or death (MALE/death) and were evaluated using Kaplan-Meier estimates and log-rank tests.

RESULTS: Of 1946 limbs undergoing a first-time intervention for CLTI between 2005 and 2022, 765 underwent BPG and 1181 underwent PTA/S. After matching, 862 fit Cohort 1 (431 BPG and 431 PTA/S), and 274 fit Cohort 2 (137 BPG and 137 PTA/S). Both cohorts exhibited a median follow-up of 2.7 years. In Cohort 1, major reintervention and MALE/death were both noted to be significantly lower following ssGSV BPG, as compared with PTA/S (at 7 years: 11% vs 24%; P = .001 and 72% vs 78%; P = .03, respectively). These findings correlated with a 53% and 28% reduction in the aforementioned adjusted events (hazard ratio, 0.47; 95% confidence interval, 0.30-0.74 and hazard ratio, 0.82; 95% confidence interval, 0.69-0.98, respectively). These significant differences in major reintervention and MALE/death were not noted in Cohort 2 (at 7 years: 25% vs 24%; P = .92 and 82% vs 80%; P = .31, respectively). Further, neither cohort demonstrated significant differences in complete wound healing (at 6 months, Cohort 1: 47% vs 40%; P = .32; Cohort 2: 40% vs 38%; P = .12), major amputation (at 7 years: Cohort 1: 15% vs 15%; P = .89; Cohort 2: 35% vs 25%; P = .86), or amputation/death (at 7 years, Cohort 1: 70% vs 66%; P = .99; Cohort 2: 78% vs 76%; P = .45).

CONCLUSIONS: Patients undergoing revascularization using ssGSV demonstrate significantly lower rates of major reintervention and MALE/death compared with those undergoing endovascular interventions for CLTI. However, similar outcomes are not seen among patients undergoing revascularization without a suitable ssGSV. These findings correlate with those demonstrated in BEST-CLI, suggesting generalizability.

OBJECTIVES: Endovascular aneurysm repair (EVAR) for large infrarenal abdominal aortic aneurysms (AAAs) has been associated with worse outcomes compared with EVAR for smaller AAAs. Whether these findings apply to complex AAAs (cAAA) remains uncertain.

METHODS: We identified all intact complex EVAR (cEVAR) from 2012 to 2024 in the Vascular Quality Initiative. cEVAR was defined as having a proximal extent between zones 6 and 9 and at least one side branch/fenestration/chimney/parallel grafting. Aneurysm size was defined as follows: large: >65 mm (males), >60 mm (females); medium: 55 to 65 mm (males), 50 to 60 mm (females); and small: <55 mm (males), <50 mm (females). We assessed perioperative death, any complication, and in-hospital reintervention using logistic regression and midterm mortality using adjusted Kaplan-Meier methods and Cox regression analyses. Medium-sized aneurysms were compared with large and small aneurysms.

RESULTS: Of the 3426 patients, 22.6% had large, 60.4% medium, and 17.0% had small aneurysms. As compared with medium and small aneurysms, large aneurysms demonstrated higher rates of perioperative death (4.8% vs 2.6% vs 0.5%), any complication (33.3% vs 23.6% vs 19.4%), and in-hospital reintervention (6.2% vs 4.0% vs 2.6%) (all P < .05). The median follow-up was 445 days. One-year mortality rates were higher in large aneurysms (12.3% vs 7.8% vs 3.8%; P < .001). After adjustment, when compared with medium-sized aneurysms, large aneurysms were associated with a significantly higher risk of perioperative death (adjusted odds ratio [aOR], 1.73; 95% confidence interval [CI], 1.09-2.72), any complication (aOR, 1.44; 95% CI, 1.18-1.76), and midterm mortality (adjusted hazard ratio, 1.50; 95% CI, 1.19-1.88), but not in-hospital reintervention (aOR, 1.46; 95% CI, 0.99-2.13). Although small aneurysms, as compared with medium-sized aneurysms, did not demonstrate a difference in any complication (aOR, 0.87; 95% CI, 0.68-1.10), in-hospital reintervention (aOR, 0.77; 95% CI, 0.42-1.33), and midterm mortality (adjusted hazard ratio, 0.78; 95% CI, 0.57-1.08], they did demonstrate a lower risk of perioperative death (aOR, 0.26; 95% CI, 0.06-0.71).

CONCLUSIONS: In cEVAR for cAAA, large aneurysms, compared with medium-sized aneurysms, were associated with higher rates of perioperative death, any complication, and midterm mortality, with in-hospital reinterventions trending toward a statistically significant higher risk. Although these results align with expectations, they emphasize the importance of effectively managing patients with large cAAAs and highlight the need for future research to determine whether patients might benefit more from medical therapy or open repair.

OBJECTIVE: Despite few clinical indications, numerous urgent or emergent carotid procedures have been recorded in the Vascular Quality Initiative (VQI) database. As such, we sought to assess outcomes of urgent and emergent carotid revascularization in the VQI.

METHODS: We identified all patients who underwent carotid revascularization in the VQI from 2011 to 2024. Patients were stratified by urgency status, preoperative symptom status, and procedure type. Elective revascularization was defined by the VQI as a planned or scheduled procedure, urgent as surgery within 24 hours, and emergent as surgery within 6 hours. We excluded patients whose primary procedure was a planned intracranial treatment, as well as patients presenting with trauma, dissection, or other nonatherosclerotic indications. The primary outcome was perioperative stroke or death. Secondary outcomes included perioperative death and stroke. We used χ2 and logistic regression to evaluate perioperative outcomes.

RESULTS: Of the 317,163 carotid revascularizations performed, 268,091 (84%) were elective, 45,021 (14%) were urgent, and 4051 (1%) were emergent. Most urgent (29,958, 67%) or emergent (2,956, 73%) cases were symptomatic, although there were 15,063 urgent (34%) and 1095 emergent (27%) among asymptomatic patients. Stroke was the indication for 44% of urgent procedures and 62% of emergent procedures. There 45,021 cases classified as urgent, of which 28,063 (62%) were carotid endarterectomy, 8172 (18%) transcarotid artery revascularization, and 8786 (19.5%) transfemoral carotid artery stenting. Of the 4051 emergent procedures, 1235 (31%) were carotid endarterectomy, 182 (4.5%) transcarotid artery revascularization, and 2634 (65%) transfemoral carotid artery stenting. Compared with elective procedures, among all patients, urgent procedures were associated with increased odds of stroke/death (3.2% vs 1.2%; adjusted odds ratio [aOR], 1.99; 95% confidence interval [CI], 1.80-2.18; P < .01), as were emergent procedures (10.4% vs 1.2%; aOR, 3.67; 95% CI, 3.03-4.44; P < .01). These differences were also noted following subset analyses of asymptomatic (urgent, 3.0% vs 1.0%; aOR, 2.52; 95% CI, 2.16-2.92; P < .01) and (emergent, 9.9% vs 1.0%; aOR, 5.5; 95% CI, 3.91-7.63; P < .01) and symptomatic patients (urgent, 3.3% vs 1.7%; aOR, 1.65; 95% CI, 1.46-1.86; P < .01) and (emergent, 11% vs 1.7%; aOR, 3.07; 95% CI, 2.43-3.86; P < .01). These differences persisted after stratifying by procedure type, for both asymptomatic and symptomatic patients.

CONCLUSIONS: Urgent or emergent carotid revascularization was associated with higher odds of perioperative stroke or death, stroke, and death. Given the increased risks of urgent or emergent surgery, careful consideration should be given when assessing patients who may ostensibly benefit from expedited surgery, where possible.

OBJECTIVE: The vast majority of patients with abdominal aortic aneurysms (AAAs) undergoing repairs receive endovascular interventions (EVARs) instead of open operations (OARs). Although EVARs have better short-term outcomes, OARs have improved longer-term durability and require less radiographic follow-up and monitoring, which may have significant implications on health care economics surrounding provision of AAA care nationally. Herein, we compared costs associated with EVAR and OAR of both infrarenal and complex AAAs.

METHODS: We examined patients undergoing index elective EVARs or OARs of infrarenal and complex AAAs in the 2014-2019 Vascular Quality Initiative-Vascular Implant Surveillance and Interventional Outcomes Network (VQI-VISION) dataset. We defined overall costs as the aggregated longitudinal costs associated with: (1) the index surgery; (2) reinterventions; and (3) imaging tests. We evaluated overall costs up to 5 years after infrarenal AAA repair and 3 years for complex AAA repair. Multivariable regressions adjusted for case-mix when evaluating cost differences between EVARs vs OARs.

RESULTS: We identified 23,746 infrarenal AAA repairs (8.7% OAR, 91% EVAR) and 2279 complex AAA repairs (69% OAR, 31% EVAR). In both cohorts, patients undergoing EVARs were more likely to be older and have more comorbidities. The cost for the index procedure for EVARs relative to OARs was lower for infrarenal AAAs ($32,440 vs $37,488; P < .01) but higher among complex AAAs ($48,870 vs $44,530; P < .01). EVARs had higher annual imaging and reintervention costs during each of the 5 postoperative years for infrarenal aneurysms and the 3 postoperative years for complex aneurysms. Among patients undergoing infrarenal AAA repairs who survived 5 years, the total 5-year cost of EVARs was similar to that of OARs ($35,858 vs $34,212; -$223 [95% confidence interval (CI), -$3042 to $2596]). For complex AAA repairs, the total cost at 3 years of EVARs was greater than OARs ($64,492 vs $42,212; +$9860 [95% CI, $5835-$13,885]). For patients receiving EVARs for complex aneurysms, physician-modified endovascular grafts had higher index procedure costs ($55,835 vs $47,064; P < .01) although similar total costs on adjusted analyses (+$1856 [95% CI, -$7997 to $11,710]; P = .70) relative to Zenith fenestrated endovascular grafts among those that were alive at 3 years.

CONCLUSIONS: Longer-term costs associated with EVARs are lower for infrarenal AAAs but higher for complex AAAs relative to OARs, driven by reintervention and imaging costs. Further analyses to characterize the financial viability of EVARs for both infrarenal and complex AAAs should evaluate hospital margins and anticipated changes in costs of devices.