Abstract
PURPOSE: The aim of our study was to determine whether delta red blood cell distribution (ΔRDW) improves neurological outcomes in acute ischemic stroke (AIS) patients 2 years after intravenous thrombolysis (IVT) therapy. METHODS: AIS patients who received IVT between January 2013 and December 2019 were retrospectively analyzed. In accordance with their mRS scores, the patients were divided into two groups. A binary logistic regression analysis was conducted to determine the influencing factors of adverse functional outcomes. It was decided to evaluate the variables' the predictive ability by using the area under the receiver operating characteristic. For the poor neurological recovery risk model, features were selected using the LASSO regression model. We also developed a predictive model based on logistic regression analysis, which combined the features selected in the minimum absolute contraction and selection operator regression models. An evaluation of the discrimination, calibration, and clinical applicability of the predictive model was conducted using the C index, calibration chart, and decision curve analysis. Internal validation was evaluated via bootstrapping. RESULTS: Binary logistic regression analysis showed that ΔRDW was an independent influencing factor for poor neurofunctional outcomes. The most appropriate ΔRDW cut-off value for predicting the recovery of poor neurological outcomes was 18.9% (sensitivity: 89.9%, specificity: 78.6%, p < 0.001). The predictive factors included in the nomogram were age, the occurrence of CHD, stroke, AF, ΔRDW, NIHSS score at onset, interval time from onset to IVT, and whether there were indwelling urine catheters and gastric tubes. The model has not only a good discrimination ability, which was indicated by an overall C index of 0.891 (95% confidence interval: 0.829-0.953), but also a considerable calibration ability. Decision curve analysis showed that the nomogram of adverse neurological outcomes recovery was useful in the clinical practice when intervention was implemented above the threshold of 1% possibility of adverse neurological outcomes recovery. CONCLUSION: In patients with AIS after thrombolysis, the ΔRDW is a potential influencing factor that can be readily used to predict the likelihood of poor neurological function recovery.