275. Clinical and Laboratory Predictors of Stroke Associated with COVID-19 Disease

Background Although SARS-CoV-2 predominantly targets the respiratory system, it has also been associated with vascular complications including stroke. Identifying COVID-19 patients at elevated risk for stroke can help inform target anticoagulation strategies. We sought to understand how symptoms and laboratory markers at presentation with COVID-19 relate to stroke risk. Methods We enrolled a cohort of 1324 subjects who were hospitalized with COVID-19 across six PennMedicine hospitals between April and August 2020 and performed retrospective, manual chart review to measure exposures including presenting symptoms and admission inflammatory markers. Data were organized with a REDCap database, and analyses were performed using R statistical software, with Bayesian binomial regression models fit using Stan Hamiltonian Monte Carlo via the “brms” package. Results Among 1324 subjects, 19 stroke events were observed within 30 days of COVID-19 diagnosis. Admission inflammatory markers, including C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), ferritin, and D-dimer, were poor predictors of stroke risk. Among presenting symptoms, including respiratory, gastrointestinal, dermatologic, and neurologic features of COVID-19 disease, only altered mental status documented on presentation (in 529 subjects) was significantly associated with stroke risk (odds ratio 6.06, 95% credible interval 2.16 - 18.7). Conclusion Inflammatory markers associated with COVID-19 disease severity did not discriminate patients at high versus low risk of stroke in this cohort. Altered mental status documented on presentation was significantly associated with incident stroke during COVID-19 disease. Disclosures Ebbing Lautenbach, MD, MPH, MSCE , Merck (Other Financial or Material Support, Member of Data and Safety Monitoring Board (DSMB)) Michael Z. David, MD PhD , GSK (Board Member)