In this cross-sectional study, ninety-five COVID-19 patients with stroke were included between October 2020 and January 2021. Patients were evaluated based on the following clinical and imaging features: severity of COVID-19 (critical/non-critical), stroke type, presence or absence of clinical suspicion of stroke (suspicious/non-suspicious), medical risk factors, Fazekas scale, ASCOD criteria classification, and presence or absence of watershed infarction. Clinical outcomes were assessed based on Modified Rankin Scale (MRS) and mortality (during hospitalization and within three months of discharge). Chi-square test, independent t-test, and one-way ANOVA were used to compare variables. Multivariable logistic regression was applied to evaluate the effects of relevant factors on the likelihood that participants have watershed infarction.

Ischemic stroke (n=79, 83.1%) and intracerebral hemorrhage (n=7, 7.4%) were the most prevalent types of stroke. According to ASCOD classification, most common causes of ischemic stroke were cardioembolic (n=21, 26.6%) or other determined causes (n=18, 22.8%). Watershed infarction was the most common etiology among other determined causes (16/18, 88.9%). Watershed infarction was significantly associated with being clinically non-suspicious (OR=8.46, p<0.001) and death after discharge (OR=12, p<0.001). Patients with watershed infarction had a higher odds of having a high Fazekas score (OR=6.42, p=0.002) which was confirmed by logistic regression model (adjusted OR=26.1, p=0.006).