Introduction: Significant demographic differences in the management of ischemic stroke and transient ischemic attack exist in the United States. Transient neurological attack (TNA) is a common neurological diagnosis that can result from cerebral ischemia, and is associated with recurrent stroke. Demographic factors may be associated with differences in admission rates for patients that present to the emergency department (ED) with a diagnosis of TNA.
Methods: In a population-based, retrospective, administrative claims-based cohort study, we identified patients who presented to a non-federal ED in Florida from 2005 through 2013 with a complaint consistent with TNA, which was defined as a poorly-localized, non-specific neurological symptom. Our predictor variables were race (white vs. non-white) and sex. Our outcome variable was admission to the hospital. We used multiple logistic regressions to characterize the relationship between the predictor variables and our chosen outcome while adjusting for age, insurance status, income, and vascular risk factors.
Results: We identified 1,231,046 patients presenting to the ED with TNA. Female patients in this cohort were less often white, and had fewer vascular comorbidities than men, including alcohol abuse. Non-white patients were younger, more often had Medicaid or no insurance, less often had atrial fibrillation, coronary heart disease or alcohol abuse, and more commonly had diabetes. Among all patients, 22,300 (1.8%) were admitted. After adjustment for age, insurance status, income, and vascular comorbidities, we found lower odds of hospital admission in female patients (odds ratio [OR], 0.69; 95% confidence interval [CI]; 0.67-0.70) and non-white patients (OR, 0.65; 95% CI, 0.63-0.67).
Conclusions: Among patients presenting to the ED with TNA, female sex and race are associated with decreased odds of admission, even after adjusting for age, income, insurance status, and vascular risk factors. Further studies are warranted to reproduce these findings in other states, and to determine the public health impact of our results.