Deconstructing health disparities: it's not just about race.

Neurology 2012;79:117–118 In view of the sacred covenant of physicians, the Hippocratic Oath, the medical community continually strives to provide timely and appropriate medical care to all patients without consideration of person, place, or insurance status. Some would purport that medical triage, acute care, and long-term follow-up in the United States are ethical, appropriate, and nonbiased, when compared with the rest of the world. However, the reality is that racial, ethnic, geographic, and social disparities continue to exist in this country.1 As the second most common cause of death worldwide, fourth cause of death in the United States, and a leading cause of adult disability, stroke is a disease with global health and economic consequences.2 When looking at race alone, black and Hispanic patients have a higher prevalence of stroke, when compared with non-Hispanic white patients.3,4 Additionally, stroke mortality is significantly higher in blacks.5 These observations clearly illustrate the importance of elucidating etiologies and interventions addressing these disparities in health.6 In this issue of Neurology, Cheng and colleagues7 examine the use of carotid artery imaging at minority-serving hospitals. Carotid artery imaging via Duplex ultrasound, CT angiography, magnetic resonance angiography, or formal cerebral angiogram represents an integral part of the standard stroke evaluation. The detection and treatment of carotid artery disease is known to reduce secondary stroke risk.8 The authors conducted a retrospective analysis of patients from a collaborative US Veterans Affairs (USVA) Special Study, which included all primary ischemic stroke admissions to the USVAMC in fiscal year 2007. Based on the patient volume of each hospital and racial distribution, medical records were systematically sampled from this database, with oversampling from high-volume facilities. There were a total of 127 hospitals included in the analysis. The final sample consisted of 1,534 white and 628 black patients. Other races/ethnicities were excluded due to small numbers. Nearly 40% of black patients were treated at 1 of 13 minority-serving hospitals. The authors defined “minority-serving” hospitals as those in the top 10th percentile by proportion of patients who were black. Statistical analyses were performed to rule out confounding variables such as age, stroke risk factors, or hospital characteristics (e.g., rural, urban, number and complexity of stroke patients) and to determine the accuracy of sampling. After controlling for significant variables, the authors found that white patients (89%) at non-minority-serving hospitals had a significantly higher probability of getting carotid imaging than either black patients (70%) or white patients (78%) at minority-serving hospitals. This study is vitally important because it illustrates a potentially modifiable health systems factor which affects not only one race of patients but all patients treated at certain facilities. When considering health disparities, most automatically think about individual patient characteristics that might lead to differences in health care delivery or outcome without regard to factors related to health systems, political-, legislative-, or advertisementrelated factors. Cheng et al. took great care to insure the accuracy of the carotid imaging data by searching the VA database 12 months before, and 2 months after, the stroke admission. The authors also conducted multiple analyses redefining minority-serving status as those with the top 10% or 25% of black patients. Despite the use of these varied parameters, the results consistently showed a significant disparity in the use of carotid imaging, when assessed by looking at minority-serving vs nonminority-serving hospital status. Strengths of this study include the analysis of a novel disparity in stroke care and a large patient population with no barrier to health care, based on insurance status or access. A limitation of this study is the lack of control for potential variability of resources at each hospital. Since the major reason for increased carotid imaging at the non-minority-serving hospitals seemed to be related to increased use of Duplex ultrasound, is it possible that the minority-serving hospitals have fewer resources and are unable to af-