Background and purposeBlood pressure variability has been found to contribute to worse outcomes after intravenous tissue plasminogen activator, but the association has not been established after intra-arterial therapies.MethodsWe retrospectively reviewed patients with an ischemic stroke treated with intra-arterial therapies from 2005 to 2015. Blood pressure variability was measured as standard deviation (SD), coefficient of variation (CV), and successive variation (SV). Ordinal logistic regression models were fitted to the outcome of the modified Rankin Scale (mRS) with univariable predictors of systolic blood pressure variability. Multivariable ordinal logistic regression models were fitted to the outcome of mRS with covariates that showed independent predictive ability (P<0.1).ResultsThere were 182 patients of mean age 63.2 years and 51.7% were female. The median admission National Institutes of Health Stroke Scalescore was 16 and 47.3% were treated with intravenous tissue plasminogen activator. In a univariable ordinal logistic regression analysis, systolic SD, CV, and SV were all significantly associated with a 1-point increase in the follow-up mRS (OR 2.30–4.38, all P<0.002). After adjusting for potential confounders, systolic SV was the best predictor of a 1-point increase in mRS at follow-up (OR 2.63–3.23, all P<0.007).ConclusionsIncreased blood pressure variability as measured by the SD, CV, and SV consistently predict worse neurologic outcomes as measured by follow-up mRS in patients with ischemic stroke treated with intra-arterial therapies. The SV is the strongest and most consistent predictor of worse outcomes at all time intervals.
Ordinal Regression Analysis: Using Generalized Ordinal Logistic Regression Models to Estimate Educational Data
