Blood viscosity values for predicting pathological stroke types (ischemic stroke, hemorrhagic stroke, and stroke mimic) at DR. Cipto Mangunkusumo National Referral Hospital.

Abstract

IntroductionStroke is the leading cause of morbidity and mortality, and its management differs among ischemic stroke, hemorrhagic stroke, and stroke mimics. Non-contrast CT has limited sensitivity for early ischemia. This study compared blood and plasma viscosity across different stroke types and identified cutoff values for predicting stroke type.MethodA cross-sectional study at the Emergency Department of Dr. Cipto Mangunkusumo National Referral Hospital (March-June 2025). Blood and plasma viscosity were measured using the Neo Digital Microcapillary device then compared using appropriate statistic tests. A p-value < 0.05 was considered statistically significant.ResultsSeventy of 97 suspect stroke patients were included (36 ischemic, 16 hemorrhagic, 18 mimics). There were significant differences of blood viscocity between ischemic and hemorrhagic stroke compared with stroke mimic (median 5.65 cP ischemic; 5.29 cP hemorrhagic; 4.84 cP mimic; p < 0.0001), whereas no significant difference was observed in plasma viscosity across the groups (p = 0.411). A blood viscosity cutoff of 5.17 cP yielded an AUC of 0.84 (p < 0.0001), 71.2% sensitivity, 94.4% specificity, 97.37% PPV, and 53.13% NPV for stroke prediction.ConclusionBlood viscosity shows potential as an adjunctive diagnostic tool for distinguishing stroke from stroke mimics with high specificity, aiding in predicting types of stroke.