AUTHOR=Peng Teng J. , Jasne Adam S. , Simonov Michael , Abdelhakim Safa , Kone Gbambele , Cheng Yee Kuang , Rethana Melissa , Tarasaria Karan , Herman Alison L. , Baker Anna D. , Yaghi Shadi , Frontera Jennifer A. , Sansing Lauren H. , Falcone Guido J. , Spudich Serena , Schindler Joseph , Sheth Kevin N. , Sharma Richa 

TITLE=Prior Stroke and Age Predict Acute Ischemic Stroke Among Hospitalized COVID-19 Patients: A Derivation and Validation Study

JOURNAL=Frontiers in Neurology

VOLUME=12

YEAR=2021

URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2021.741044

DOI=10.3389/fneur.2021.741044

ISSN=1664-2295

ABSTRACT=<p><bold>Objectives:</bold> Our objective was to identify characteristics associated with having an acute ischemic stroke (AIS) among hospitalized COVID-19 patients and the subset of these patients with a neurologic symptom.</p><p><bold>Materials and Methods:</bold> Our derivation cohort consisted of COVID-19 patients admitted to Yale-New Haven Health between January 3, 2020 and August 28, 2020 with and without AIS. We also studied a sub-cohort of hospitalized COVID-19 patients demonstrating a neurologic symptom with and without an AIS. Demographic, clinical, and laboratory results were compared between AIS and non-AIS patients in the full COVID-19 cohort and in the sub-cohort of COVID-19 patients with a neurologic symptom. Multivariable logistic regression models were built to predict ischemic stroke risk in these two COVID-19 cohorts. These 2 models were externally validated in COVID-19 patients hospitalized at a major health system in New York. We then compared the distribution of the resulting predictors in a non-COVID ischemic stroke control cohort.</p><p><bold>Results:</bold> A total of 1,827 patients were included in the derivation cohort (AIS <italic>N</italic> = 44; no AIS <italic>N</italic> = 1,783). Among all hospitalized COVID-19 patients, history of prior stroke and platelet count ≥ 200 × 1,000/μL at hospital presentation were independent predictors of AIS (derivation AUC 0.89, validation AUC 0.82), irrespective of COVID-19 severity. Among hospitalized COVID-19 patients with a neurologic symptom (<italic>N</italic> = 827), the risk of AIS was significantly higher among patients with a history of prior stroke and age &lt;60 (derivation AUC 0.83, validation AUC 0.81). Notably, in a non-COVID ischemic stroke control cohort (<italic>N</italic> = 168), AIS patients were significantly older and less likely to have had a prior stroke, demonstrating the uniqueness of AIS patients with COVID-19.</p><p><bold>Conclusions:</bold> Hospitalized COVID-19 patients who demonstrate a neurologic symptom and have either a history of prior stroke or are of younger age are at higher risk of ischemic stroke.</p>