Research On Prediction Efficiency Of Modified WORSEN Score Based On White Matter Lesions For Early Neurological Deterioration In Acute Ischemic Stroke

Objective: Investigating the correlation between white matter lesions and Early Neurological Deterioration(END)in acute ischemic stroke;Evaluating the predictive efficacy of WORSEN score and Fazekas score for END in acute ischemic stroke;We improved the WORSEN score based on the Fazekas score,assessed its prediction efficiency for END in patients with acute ischemic stroke,and to assist clinicians to judge the early prognosis of patients.Materials and Methods: From September 2019 to November 2021,260 patients who met the admission and exclusion criteria and were admitted to the neurology department of The First Affiliated Hospital of Dalian Medical University Hospital were prospectively included.General data,laboratory data and imaging data of all patients were collected.In this study,END was defined as acute ischemic stroke patients admitted within 48 hours of the onset of the disease presenting worsening neurological impairment symptoms within 7 days and an increase of ≥2 points in NIHSS score.According to the occurrence of END,the patients were divided into END group and non-END group and understand the incidence of END.Univariate analysis was conducted on the collected variables to analyze whether they were risk factors for the occurrence of END.Fazekas scores were performed in the enrolled patients,and Pearson correlation was used to analyze the correlation between Fazekas scores of white matter lesions and the incidence of END.The WORSEN score was applied to the enrolled patients,and the ROC curve of the WORSEN score for predicting END was plotted to evaluate its predictive efficiency for END in acute ischemic stroke.The ROC curve of Fazekas score for predicting END was plotted to evaluate its predictive efficacy for END in acute ischemic stroke.The WORSEN score was improved based on the Fazekas score for white matter lesions.Evaluating the predictive efficacy of the modified score for END and comparing the statistical difference between the WOSREN score and the improved score in the predictive effect of END.Results:1.A total of 260 cases of acute ischemic stroke were enrolled,including 161males(61.9%)and 99 females(38.1%),and the mean age was 65.47±10.89.Early neurological deterioration occurred in 63 patients of 260 patients(24.23%).2.Univariate analysis showed that the baseline NIHSS score,fasting blood glucose level,glycosylated hemoglobin level,occluding rate of internal carotid artery,M1 segment occlusion of middle cerebral artery and WORSEN score were higher in the END group than in the non-END group,and they were significantly different(P<0.05).3.Pearson correlation analysis was conducted between the incidence of END and Fazekas score.The Fazekas score was positively correlated with the incidence of END,and the difference was statistically significant(r=0.963,P=0.001).With the increase of Fazekas score,the incidence of END gradually increased.4.The area under ROC curve of the WORSEN score for predicting END in patients with acute ischemic stroke was 0.785(95%CI: 0.730-0.833,P=0.000),the Youden index was 0.4287,the cut-off value of the WORSEN score was 3 points,the sensitivity was 68.25%,and the specificity was 74.62%.The area under ROC curve of Fazekas score for predicting END in patients with acute ischemic stroke was 0.741(95%CI: 0.683-0.793,P=0.000),the Youden index was 0.3457,the cut-off value of Fazekas score was 2 points,the sensitivity was 88.89%,and the specificity was45.69%.5.We improved the WORSEN score based on the Fazekas score of white matter lesions.The AUC values of the eight modified scores in predicting END were higher than the original WORSEN scores,and the predictive ability was better.The modified score 2 was the WORSEN score plus the Fazekas score(≤3 points of 0,> 3 points of 1point)and its AUC for predicting END was 0.819(95%CI:0.767～0.864,P=0.000),a cut-off value of 3 points,a sensitivity of 76.19%,and a specificity of 74.62%.The modified score 8 was the WORSEN score plus the Fazekas score(≤4 points of 0,> 4points of 1 point)and its AUC for predicting END was 0.801(95%CI: 0.747-0.847,P=0.000),a cut-off value of 3 points,a sensitivity of 71.43%,and a specificity of74.62%.The sensitivity of the above two modified scores in predicting END was higher than the original WORSEN score,and the AUC value of predicting END was higher than the original WORSEN score,and the difference was statistically significant(P=0.0021,P=0.0198).Conclusions:1.The incidence of END in patients was 24.23%,and the occurrence of END was related to baseline NIHSS level,fasting blood glucose,glycosylated hemoglobin,internal carotid artery occlusion,and M1 segment occlusion of middle cerebral artery.2.The severity of white matter lesions is positively correlated with the incidence of END.The more severe the severity of white matter lesions,the higher the incidence of END in patients with ischemic stroke.3.The WORSEN score and Fazekas score have good predictive efficacy for END.The modified scores based on the Fazekas score of white matter lesions have a better predictive efficacy for END.At the same time,they can improve the sensitivity of predicting END.Among them,the modified score 2,namely the WORSEN score+Fazekas score(≤3 points of 0,> 3 points of 1),has the best predictive efficacy for END,which is helpful for clinical prediction of the occurrence of END.