Endovascular aneurysm repair conversion is an increasingly common indication for open abdominal aortic aneurysm repair

Although endovascular aneurysm repair (EVAR) reintervention is common, conversion to open repair (EVAR-c) occurs less frequently but can be associated with significant technical complexity and perioperative risk. There is a paucity of data highlighting the evolution of periprocedural results surrounding EVAR-c and change in practice patterns, especially for referral centers that increasingly manage EVAR failures. The purpose of this analysis was to perform a temporal analysis of our EVAR-c experience and describe changes in patient selection, operative details, and outcomes. A retrospective single-center review of all open abdominal aortic aneurysm repairs was performed (2002-2019), and EVAR-c procedures were subsequently analyzed. EVAR-c patients (n = 184) were categorized into two different eras (2002-2009, n = 21; 2010-2019, n = 163) for comparison. Logistic regression and Cox proportional hazards modeling were used for risk-adjusted comparisons. A significant increase in EVAR-c as an indication for any type of open aneurysm repair was detected (9% to 27%; P < .001). Among EVAR-c patients, no change in age or individual comorbidities was evident (mean age, 71 ± 9 years); however, the proportion of female patients (P = .01) and American Society of Anesthesiologists classification >3 declined (P = .05). There was no difference in prevalence (50% vs 43%; P = .6) or number (median, 1.5 [interquartile range (IQR), 0-5]) of preadmission EVAR reinterventions; however, time to reintervention decreased (median, 23 [IQR, 6-34] months vs 0 [IQR, 0-22] months; P = .005). In contrast, time to EVAR-c significantly increased (median, 16 [IQR, 9-39] months vs 48 [IQR, 20-83] months; P = .008). No difference in frequency of nonelective presentation (mean, 52%; P = .9] or indication was identified, but a trend toward increasing mycotic EVAR-c was observed (5% vs 15%; P = .09). Use of retroperitoneal exposure (14% vs 77%; P < .0001), suprarenal cross-clamp application (6286%; P = .04), and visceral-ischemia time (median, 0 [IQR, 0-11] minutes vs 5 [IQR, 0-20] minutes; P = .05) all increased. In contrast, estimated blood loss (P trend = .03) and procedure time (P = .008) decreased. The unadjusted elective 30-day mortality rate improved but did not reach statistical significance (elective, 10% vs 5%; P = .5) with no change for non-elective operations (18% vs 16%; P = .9). However, a significantly decreased risk of complications was evident (odds ratio, 0.88; 95% confidence