Abstract W P304: Variation in Use of Carotid Revascularization Procedures and In-Hospital Mortality Among Large, Urban Teaching Hospitals in the United States

Background: Guidelines and quality improvement efforts seek to minimize variation in care and outcomes across hospitals. We assessed hospital-level variation in procedure rates and in-hospital mortality for patients hospitalized with ischemic stroke at similar hospitals in the US. The use of procedures and in-hospital mortality were not expected to vary significantly among comparable, high-volume facilities after adjusting for patient case-mix. Methods: We selected urban teaching hospitals with ≥100 annual ischemic stroke discharges (ICD-9 433, 434, 436) from the Nationwide Inpatient Sample 2010-2011. Generalized linear mixed models were used to quantify between-hospital variation in the use of carotid artery stenting (CAS) and endarterectomy (CEA), as well as in-hospital mortality, adjusting for patient characteristics. Adjusted odds ratios were calculated to reflect the odds that patients would have the procedure/outcome when treated at hospitals 1 SD above relative to hospitals 1 SD below the overall rate for that procedure/outcome (an odds ratio of 1.0 would reflect no hospital variation in the procedure/outcome). Results: A total of 105 urban teaching hospitals were selected, with a median annual volume of 453 ischemic stroke discharges (IQR 351-600). Among a total of 52,090 ischemic stroke discharges (mean age 68±14.8 yrs), the overall rates were 3.7% (SD 3.1) for CAS and 15.6% (SD 8.0) for CEA; in-hospital mortality was 4.3% (SD 1.7). The odds of receiving CAS and CEA were almost 7 and 4 times as high, respectively, for a patient treated at a hospital 1 SD above versus 1 SD below the overall rate for that procedure (CAS: 6.68, 95% CI 4.97-8.98; CEA: 3.62, 95% CI 3.17-4.13). The odds of dying for those treated at a hospital 1 SD above relative to 1 SD below the overall mortality rate were 2.09 (95% CI 1.98-2.21). Conclusions: There was marked between-hospital heterogeneity in the use of carotid revascularization procedures and in-hospital mortality among large, urban US teaching hospitals. Future research needs to identify system-level factors contributing to these variations in care and outcomes.