SARS-CoV-2 infection increases risk of intracranial hemorrhage

We examined 21 patients suffering from stroke and intracranial hemorrhage (ICH) and 9 (43%) of them were male. We compared relative frequencies using Fisher’s exact test. As we had few observations and many variables, we used principal component analysis (PCA) to reduce data dimensionality. We trained a linear support vector machine (SVM) on the first two PCs of the laboratory data to predict COVID-19.ResultsPatients suffering from stroke had either hypertension or SARS-CoV-2 infection, but seldom both (OR = 0.05, p = 0.0075). The presence of SARS-CoV-2 infection was strongly associated with the logarithm of CRP (p = 1.4e–07) and with D-DIMER (p = 1.6e–05) and moderately with PT (p = 0.0024). SARS-CoV-2 infection was not related to any other factor. CRP, D-DIMER, PT, and INR were all related to each other (R2 ranging from 0.19 to 0.52, p ranging from 0.012 to < 0.0001). The first two PCs covered 96% of the variance in the four variables. Using them, perfect linear discrimination between patients suffering from COVID-19 and other patients could be achieved.Discussion and conclusionSARS-CoV-2 infection causes systemic inflammation, which is suggested as a predictor of the severe course of ICH. SARS-CoV-2 infection is an additional risk factor for vascular complications.
Source: Frontiers in Human Neuroscience - Category: Neuroscience Source Type: research