Prognostic models for ambulatory recovery functions in acute ischemic stroke patients with thrombolysis therapy

Abstract

Background. The purpose of this study was to develop models to predict the recovery of ambulatory functions taking into account the capability of the motor system to functionally reorganize in response to thrombolysis therapy. Methods. We predicted ambulatory functions recovery using retrospective data from a stroke registry of acute ischemic stroke patients who received thrombolysis therapy. Multivariate regression was used to construct the models. Multicollinearity and significant interactions were examined using variance inflation factors, while a Cox & Snell classification were applied to check the fitness of each model. Results. The models correctly predicted clinical variables that were associated with an improvement or non-improvement in functional ambulatory outcome. Broca’s aphasia (OR = 2.270, P = 0.002, CI =1.34-3.83) was associated with improved functional outcome at discharge, while patients aged 80 years or older (OR = 0.942, P = <0.001, CI =0.92-0.96), patients with congestive heart failure (OR = 0.496, P = 0.040, CI = 0.25-0.97), higher NIHSS (OR=0.876, P = 0.001, CI = 0.80-0.95), taking antihypertensive medication (OR = 0.436, P = 0.023, CI = 0.21-0.89) were not associated with improved ambulatory functional outcome with thrombolysis. The discriminating ability for the model was 74.2% for the total population, 71.7% for the rtPA group, and 72.2% for the no-rtPA group indicating strong performance. Conclusion. Prognostic models that can predict improved functional ambulatory outcome in thrombolysis therapy can be beneficial in the care of stroke patients. Our models predicted improved functional recovery of Broca’s aphasia after thrombolysis therapy, suggesting a future potential to evaluate motor speech area after stroke.