Abstract 13: Disparities in Carotid Imaging Among Persons Presenting with Acute Ischemic Stroke

Abstract

Background: Atherosclerosis from internal carotid artery stenosis is a major cause of stroke. It is identified through imaging of the carotid arteries, which should be routinely performed in the diagnostic workup for acute ischemic stroke. In prior research, we identified black race as an independent predictor of the underuse of carotid imaging. We explored factors that may explain this racial disparity. Methods: Data were obtained from a retrospective study of 3965 persons admitted to 129 Veterans Administration Medical Centers (VAMC) with ischemic stroke in 2007. To identify a sample who should definitely receive imaging, extensive exclusion criteria were applied such as transfers, inpatient deaths, and high levels of comorbidity. Among 1432 whites versus 610 blacks, we compared receipt of carotid artery imaging between 12 months before admission to 2 months after discharge in bivariate and multivariate analyses. We then compared patient characteristics, hospital characteristics, and the distribution of underuse and disparities across VAMC among whites versus blacks to identify explanations for the racial disparity. Results: Blacks were less likely to receive carotid imaging than whites (81% vs. 89%, p<0.001). In multivariate hierarchical models, blacks remained less likely to receive carotid imaging (odds ratio 0.67, 95% confidence interval [0.48 to 0.93] after adjustment for age, presence of atrial fibrillation, Charlson Comorbidity Index, and NIH Stroke Scale, suggesting that clinical factors were not the reason for the disparity. Black patients were more likely to attend larger, higher complexity, urban hospitals, suggesting that hospital characteristics were not responsible for the disparity. The racial disparities at the VAMC level were greatest among those that had the lowest use of carotid imaging. Two VAMCs accounted for 25% of the system-wide magnitude of disparity in the VA. When these two VAMCs were removed from the sample, race was no longer an independent predictor of underuse. Conclusions: Racial disparities of carotid imaging were not mediated by other patient or hospital characteristics. Improving overall receipt of carotid imaging at a few VAMCs can substantially reduce the system-wide magnitude of disparities in the VA.