Abstract TP294: Women Admitted for Suspected TIA are Less Likely to Have Imaging Evidence of Infarction than Men: Results from the Rhode Island Hospital Prospective TIA unit cohort

Abstract

Introduction: Previous studies have shown sex differences in the presentation and management of patients with transient ischemic attack (TIA), but little is known about how outcomes differ by sex among those admitted to emergency department observation units (ED OU), an increasingly common ED disposition for patients with suspected TIA. Objective: To determine if there are sex differences in the rate of positive diffusion weighted MRIs (DWI) among patients with suspected TIA admitted to an ED OU. Methods: Patients in a large, urban, academic ED admitted to the ED OU for suspected TIA from 4/2013 - 7/2018 were included. Patients with persistent deficits, fever, heart rate <60 / >100, SBP >180 / < 100 mm Hg, pulse ox <93%, or other competing diagnoses were excluded. Standard blood tests, EKG, echocardiogram, MRI, and neurology consultation were performed. Rates of acute infarct on DWI were compared between women and men in unadjusted analyses, followed by multivariable logistic regression. The final model included covariates that were significantly associated with infarct on DWI in unadjusted analyses (p<0.05). Results: 1208 patients were included; (52.9% women, 24.3% non-white). Women and men were of similar age (63.4 vs. 64.8) and had similar median duration of symptoms (45: IQR (15-90) vs 30 (10-90) min, p=0.51). Less women than men had hypertension (59.0% vs. 66.6%, p=0.02) or diabetes (17.5% vs. 22.5%, p=0.01), while more women had histories of migraines (12.8% vs. 3.5%, p<0.001). More women than men had pain on presentation (30.5% vs. 21.4%, p=0.001) and had a discharge diagnosis of something other than TIA/stroke (45.3% vs. 35.5%, p=0.002). Unadjusted, 19.1% vs. 13.0% had acute infarcts on DWI (p=0.08). After adjusting for age, race, history of hypertension, prior stroke/ TIA, and presenting symptoms (TABLE), women were less likely than men to have infarcts on DWI (aOR 0.55, 95% CI 0.38-0.79, p=0.001). Conclusions: Among patients with suspected TIA admitted to an ED OU, women were less likely to have acute infarcts on DWI. Our findings of sex differences in DWI infarct rate as well as co-morbidities and presenting symptoms may suggest a sex difference in diagnostic uncertainty and/or stroke mimics among those with suspected TIA/mild stroke.