Occurrence of “never events” after major open vascular surgery procedures

Objective “Never events” refers to harmful hospital-acquired conditions that the Centers for Medicare and Medicaid Services identified in 2008 as largely preventable and that would no longer be reimbursed. Our goal was to identify the incidence, predictive factors, temporal trend, and associated consequences of never events after major open vascular surgery procedures. Methods The Nationwide Inpatient Sample (NIS) (2003-2011) was queried to identify never events applicable to vascular surgery patients, including air embolism, catheter-based urinary tract infections (UTIs), stage 3 and 4 pressure ulcers, falls/trauma, blood incompatibility, vascular catheter infections, complications of poor glucose control, retained foreign objects, and wrong-site surgery. We specifically evaluated open abdominal aortic aneurysm repair, carotid endarterectomy, and lower extremity bypass/femoral endarterectomy. Multivariable logistic regression was used to predict never events based on preoperative variables. Multivariable logistic and gamma regression models were used to study mortality, hospital length of stay (LOS), and charges. Results Never events were identified in 774 of 267,734 patients. The distribution of never events were falls/trauma (59%), pressure ulcers (19%), catheter-based UTI (9%), vascular catheter infection (6%), complications of poor glucose control (5%), and retained objects (4%). Rates of falls and catheter-based UTIs have increased since 2008. Multivariable predictors of any never event included lower extremity bypass, abdominal aortic aneurysm, weight loss, nonelective admission, paralysis, repair, congestive heart failure, altered mental status, renal failure, weekend admission, diabetes, female gender, and age. Race, insurance, hospital type, income level, geography, July to September admission, and other comorbidities were not predictive. After risk factor adjustment, never events were associated with increased perioperative mortality (odds ratio, 2.7; 95% confidence interval [CI], 1.5-34.8; P < .001), LOS (means ratio, 1.9; 95% CI, 1.7-2.0; P