Regional Variation in Health Care Utilization and Outcomes in Ischemic Stroke Linda S. Williams, MD,*,†,‡ George J. Eckert, MAS,§ Gilbert J. L’italien, PhD, Pablo Lapuerta, MD, and Morris Weinberger, PhD*,‡,§ Little information is available about regional variation in health care utilization and the effect of utilization on outcomes after ischemic stroke. The goal of this study was to investigate the effect of regional variation in outpatient health care use on mortality after ischemic stroke. We performed a retrospective cohort study of 55,094 veterans hospital- ized for ischemic stroke at any US Veterans Affairs Medical Center between October 1, 1990, and September 30, 1997. We extracted administrative data on patient demographics, coexisting medical conditions, site of hospitalization, inpatient and outpatient health care utilization, and all-cause mortality during hospitalization and after stroke discharge. Predictors of long-term mortality in patients surviving at least 60 days post-stroke were modeled using Cox regression. Patients in the Northeast part of the country had higher comorbidity scores, a longer median length of stay, and higher in-hospital mortality than patients in other regions. However, Northeast and West patients had lower all-cause mortality after stroke than those in the Midwest or South. Patients in the Northeast (28%) and West (32%) were also more likely than those in the South (21%) or Midwest (22%) to have a neurology and/or general medicine visit within 60 days of discharge (P  .001). Adjusted mortality (HR, 95% CI) was lower in the Northeast (0.84, 0.80-0.88) and West (0.93, 95% CI 0.89, 0.97), and in patients with neurology (0.72, 0.67-0.77) or general medicine (0.85, 0.81-0.89) follow-up within 60 days of stroke discharge. We concluded that regional variation exists in patient outcomes and patterns of care following stroke. Mortality is lower in regions where more patients have early outpa- tient care after stroke. Prospective studies evaluating the cause and impact of these variations are needed to identify optimal stroke care practices. Copyright © 2003 by National Stroke Association Geographic variation in US hospital use and patient outcomes is common. Even in the health care system of the Department of Veterans Affairs (VA), which has centralized administration, more homogeneous patient characteristics, and salaried physicians, significant re- gional variations have been reported for some chronic diseases. 1 Whether similar regional variation is evident in ischemic stroke-related utilization and outcomes is un- known. Examining regional variation in stroke-related care can help identify patient and process characteristics that influence stroke outcomes and, ultimately, strategies that may optimize inpatient and outpatient stroke care. The objective of this study was to describe regional patterns of patient utilization and mortality for a national cohort of veterans with ischemic stroke. Methods We used VA administrative databases to construct a national cohort of patients discharged from any US VA Medical Center (VAMC) with a primary diagnosis of From the *Roudebush VAMC, Health Services Research & Devel- opment, the †Department of Neurology, Indiana University School of Medicine, the ‡Regenstrief Institute for Health Care, and the §Department of Medicine, Indiana University School of Medicine, Indianapolis, Indiana; and the Pharmaceutical Research Institute, Bristol Myers Squibb, Inc., New York, New York. Received August 1, 2003; accepted September 3, 2003. Supported by an Advanced Research Career Development Award, Department of Veterans Affairs, Health Services Research and De- velopment (L.S.W) and a contract from Bristol Myers Squibb, Inc. It was performed, in part, in the Regenstrief Institute for Health Care. Address reprint requests to Linda S. Williams, MD, Roudebush VAMC, HSR&D 11H, 1481 West 10th Street, Indianapolis, IN 46202. E-mail: lwilliams@hsrd.va.iupui.edu Copyright © 2003 by National Stroke Association 1052-3057/03/1206-0002$30.00/0 doi:10.1016/j.jstrokecerebrovasdis.2003.09.008 Journal of Stroke and Cerebrovascular Diseases, Vol. 12, No. 6 (November-December), 2003: pp 259-265 259