Real-world evidence of superiority of endovascular repair in treating ruptured abdominal aortic aneurysm

Abstract

Objective: The majority of previous studies, including randomized controlled trials, have failed to provide sufficient evidence of superiority of endovascular aneurysm repair (EVAR) over open aortic repair (OAR) of ruptured abdominal aortic aneurysm (rAAA) while comparing mortality and complications. This is in part due to small study size, patient selection bias, scarce adjustment for essential variables, single insurance type, or selection of only older patients. This study aimed to provide real-world, contemporary, comprehensive, and robust evidence on mortality of EVAR vs OAR of rAAA.

Methods: A retrospective observational cohort study was performed of rAAA patients registered in the Premier Healthcare Database between July 2009 and March 2015. A multivariate logistic regression model was operated to estimate the association between procedure types (OAR vs EVAR) and in-hospital mortality. The final model was adjusted for demographics (age, sex, race, marital status, and geographic region), hospital characteristics (urban or rural, teaching or not), and potential confounders (hypertension, diabetes, hypercholesterolemia, obesity, ischemic heart disease, chronic kidney disease, symptoms of critical limb ischemia, chronic obstructive pulmonary disease, smoking, and alcoholism). Furthermore, coarsened exact matching was applied to substantiate the result in the matched cohort.

Results: There were a total of 3164 patients with rAAA (1550 [49.0%] OAR and 1614 [51.0%] EVAR). Mortality was 23.79% in the EVAR group compared with 36.26% in the OAR group (P < .001). The adjusted odds ratios of mortality (1.91; 95% confidence interval [CI], 1.62-2.25; P < .001), cardiac complication (1.54; 95% CI, 1.22-1.96; P < .001), pulmonary failure (1.90; 95% CI, 1.60-2.24; P < .001), renal failure (1.90; 95% CI, 1.61-2.23; P < .001), and bowel ischemia (2.40; 95% CI, 1.70-3.35; P < .001) were significantly higher after OAR compared with EVAR. We further applied coarsened exact matching, which followed the same pattern of mortality (odds ratio, 1.68; 95% CI 1.41-1.99; P < .001) and all major complications.

Conclusions: Although the choice of repair of rAAA is highly dependent on the experience of the operating team and the anatomic suitability of the patient, this contemporary analysis of a large cohort of rAAA showed significantly higher adjusted risk of mortality in OAR compared with EVAR and substantially higher complications.