Mortality Is Not Associated With Type Of Anesthesia In Patients Undergoing Carotid Endarterectomy Or Carotid Artery Stenting; Differences In Large Database Outcomes
Alexander Ostapenko, MD, Stephanie Stroever, PhD, MPH, Alan Dietzek, MD.
Danbury Hospital, Danbury, CT, USA.
OBJECTIVES: Previous published studies have failed to demonstrate a difference in mortality for carotid endarterectomy (CEA) performed under general versus local anesthesia. Alternatively, a recent publication utilizing the Vascular Quality Initiative (VQI) reported an increased mortality in patients undergoing carotid artery stenting (CAS) with general anesthesia (GA) vs local anesthesia (LA). Utilizing the National Surgical Quality Improvement Program (NSQIP) database, we sought to replicate the results from VQI in patients undergoing CAS, extend our analysis to patients undergoing CEA, and qualitatively assess differences and similarities between the two databases.
METHODS: We conducteda cross-sectional retrospective analysis of the targeted CAS and CEA-NSQIP database for 2011 to 2017 with a primary outcome of death. We performed a univariate logistic regression analysis for the baseline characteristics and primary outcome of interest. We utilized a logistic multivariate regression model to identify variables predictive of the outcomes of interest. We then qualitatively compared the results from our NSQIP analysis to the results of the VQI study.RESULTS: A total of 23,381 patients were included in the study: 22,512 underwent CEA and 869 CAS. In the CEA group, 20,105 (89%) had GA and 2,407 (10.7%) had LA. Conversely in the CAS group 270 (31%) had GA and 599 (69%) LA. On multivariate analysis, the type of anesthesia was not a significant predictor of death (p=0.22) in the CEA group, while factors including age (p<0.001), smoking (p=0.028), high ASA class (p=0.0036), and history of congestive heart failure (p<0.001) were. In the CAS group, death was associated with anesthesia type on univariate analysis (p=0.031) but after controlling for confounders in our multivariate analysis we demonstrated anesthesia type was no longer a significant predictor of mortality.
CONCLUSIONS: Our study utilizing NSQIP database shows that anesthesia type is not associated with increased mortality in patients undergoing either CAS or CEA. These findings are in contrast to those found in a recent study utilizing VQI dataset. Both VQI and NSQIP databases are useful in assessing vascular patients among specific populations, but further research is necessary when large databases yield different results. With an increasing number of publications reporting outcomes based on large databases, our study highlights the need to take pause before assuming that any one database can provide definitive answers to all questions.
Back to 2020 ePosters