Predictors affecting in-hospital mortality of ruptured abdominal aortic aneurysms: a Greek multicenter study.

Background Endovascular aortic repair (EVAR) is being used with increasing frequency for the treatment of ruptured abdominal aortic aneurysms (rAAAs), although conflicting results have been reported concerning perioperative mortality. The aim of our study was to evaluate potential difference in mortality rates between EVAR and open surgical repair (OSR) and identify independent risk factors for in-hospital mortality in rAAAs. This study also aimed to evaluate the Glasgow Aneurysm Score (GAS) in predicting in-hospital mortality. A time-trend analysis of EVAR for ruptured AAAs was also performed. Methods Prospectively collected data from 7 public hospitals in Greece concerning rAAA repairs between January 2006 and April 2012 were analyzed. Primary outcome was in-hospital mortality. Multivariate logistic regression analysis was used to identify independent risk factors. The receiver–operator characteristic curve was used to determine the value of the GAS in predicting in-hospital death. Time-trend analysis, depicting annual changes (%), concerning EVAR for ruptured AAAs was also conducted. Results A total of 418 patients (92.3% men, mean age = 74.3 ± 8.8) with rAAAs were recorded during the study period. Among them, 113 patients (27%) underwent EVAR. Overall in-hospital mortality was 45.2%, whereas in-hospital mortality after EVAR and OSR was 20.4% and 54.3%, respectively ( P P P P P P  = 0.02) independently increased in-hospital mortality. Area under the curve of GAS was 0.80 (95% confidence interval [CI] = 0.75–0.85, P P  = 0.04) for EVAR. Annual increase of proportion (%) of EVAR for rupture was 5% ( P  = 0.004). Conclusions EVAR is being used with increasing frequency for the treatment of rAAAs and it appears to be associated with lower in-hospital mortality compared with OSR, after adjustment for hemodynamic instability and known atherosclerotic risk factors. Preoperative predictors of in-hospital mortality such as GAS should be probably modified in these patients.