IN-HOSPITAL MORTALITY FOLLOWING CAROTID ENDARTERECTOMY

Objectives: To determine population-based estimates of in-hospital mor tality following carotid endarterectomy (CEA) and identify potential r isk factors for in-hospital death. Methods: Data from the Healthcare C ost and Utilization Project (HCUP-3) were analyzed for the year 1993. Nationally representative estimates of risk were calculated by age, se x, race, income, census region, hospital location (urban versus rural) , teaching status of hospital, number of hospital beds, hospital owner ship, third-party payer, principal procedure, and presence of surgical complications. Multivariate models were developed using stepwise logi stic regression and a legit model fit by generalized estimating equati ons. Results: There were 228 deaths among 18,510 CEAs performed in 17 states of the United States in 1993, yielding an estimated in-hospital mortality rate of 1.2%. Multivariate analysis showed that age, princi pal procedure, and presence of any surgical complication were signific ant predictors of in-hospital mortality. Mortality increased with incr easing age (from 0.9% in those younger than 65 years to 1.7% in those age 75 and older) and was markedly higher with CEA performed as a seco ndary procedure (6.1% versus 0.9%) or with any surgical complication ( 5.9% versus 0.9%). Conclusions: Increasing age, CEA performed as a sec ondary procedure, and surgical complications are important predictors of in-hospital mortality following CEA.