Goldstein, Eric

Background: It is known that there are discrepancies between insurance status and access to care for a variety of health conditions. In addition, differences have been shown in initial in-hospital management of stroke as well as outcomes in the acute setting. We sought to examine the relationship between insurance status and risk of recurrent stroke, major bleeding or death following an incident cryptogenic stroke. Methods: Patients with cryptogenic stroke included in a multicenter (n=27 sites) retrospective observational cohort and complete insurance status were evaluated for a primary composite outcome of recurrent stroke, major bleeding, or death. Insurance providers were categorized by Private/Veterans Administration, Medicare, Other, and Medicaid/pending/None, with the primary carrier ranked in that order when multiple insurance carriers applied. Descriptive statistics and an adjusted Cox proportional hazards regression were used to associate the primary insurance provider with the composite outcome, and with each unique secondary outcome, adjusting for age, stroke severity, pre-stroke disability, atherosclerotic risk factors, and potential embolic sources (e.g., patent foramen ovale). Results: Of the 2201 patients, 2006 (91.1%) had reported insurance information and were followed for a median of 1.50 years (interquartile range [IQR 0.41 - 2.99y]). The majority of patients were primary Medicare beneficiaries (42.3%), followed by private insurance (36.7%), Medicaid/none (18.1%) and Other (2.9%). Although there were important differences in demographic and neurodiagnostic results across insurance groups, compared to Private/VA beneficiaries, beneficiaries for Medicaid/pending/none had a higher rate of the composite outcome in adjusted regression (adjusted hazard ratio [HR] 1.35, 95% confidence interval [CI], 1.07-1.69) which was driven largely by the higher rate of recurrent stroke (adjusted HR 1.33, 95%CI, 1.02-1.73) with no significant difference in rate of major bleeding or death (p>0.05 for both). Conclusions: Independent of age, race, pre-existing disability, and atherosclerotic risk factors, insurance status was independently predictive of recurrent stroke in this diverse cohort of cryptogenic stroke.

Introduction: Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is a genetic, progressive microangiopathy caused by mutations in the NOTCH3 gene. Patients often present with migraine and progressive subcortical infarcts resulting in vascular cognitive impairment and death. Currently, there are no disease modifying treatments. Limiting future trials is the lack of an objective construct to classify and monitor severity of disease. The purpose of this study was to gain experience with a novel CADASIL severity grading system. Methods: Neurologists at 5 academic centers retrospectively reviewed charts of those with confirmed CADASIL. Demographic and clinical data were recorded. The CADASIL severity grades were then assigned: 0 (asymptomatic), 1 (migraine alone), 2 (stroke or mild cognitive impairment), 3 (gait impairment or early dementia) and 4 (bedbound) based on available data. Standard descriptive statistical analyses were used. Results: We identified 138 patients; overall mean age of 50.9±13.1 years, 42.8% were men. 15 (10.9%) were grade 0, 50 (36.2%) were grade 1, 41 (44.2%) were grade 2, 12 (8.7%) were grade 3, and none were grade 4. Those with less severe disease tended to be younger, more likely men, had a lower rate of hypertension and diabetes mellitus and were less frequently tobacco users, Table 1. Increasing number of vascular risk factors was associated with higher severity grading. Conclusion: Our study highlights the ability of this severity scale to separate those with CADASIL into an ordinal system. As known in the literature, more advanced phenotypes of CADASIL tend to have more vascular risk factors. Future trials may consider utilizing this ordinal construct to monitor disease progression, response to therapy, or as a screening tool.