摘要
目的探讨ASTRAL量表评分对后循环急性脑梗死（ACI）患者预后的评估价值。方法选择2011年6月—2014年6月于华北理工大学附属医院神经内科住院的经颅脑MRI明确诊断的后循环ACI患者211例为研究对象。患者在入院24 h内采用ASTRAL量表进行评分,发病30 d时采用改良Rankin量表（mRS）评分进行预后评估。按照mRS评分将患者分为预后良好组和预后不良组。描绘ASTRAL量表评分评估后循环ACI患者预后的受试者工作特征曲线（ROC曲线）,计算ROC曲线下面积（AUC）、最大Youden指数,找出理想界值;以预后为因变量,ASTRAL量表评分为自变量,进行Fisher法线性判别分析。结果预后良好组167例,预后不良组44例。预后良好组患者ASTRAL量表评分低于预后不良组（t=-7.66,P〈0.05）。预后不良率随ASTRAL量表评分增加而升高（P〈0.05）。ASTRAL量表评分评估后循环ACI患者预后的AUC为0.788,95%CI（0.712,0.864）,理想界值为19分,灵敏度为72.7%,特异度为68.9%,正确率为69.7%。按照理想界值将患者分为高分组（≥19分,84例）和低分组（〈19分,127例）。高分组患者预后不良率高于低分组（χ~2=25.14,P〈0.01）。Fisher法线性判别分析得到判别方程式：Y_0=-6.807＋0.709X,Y_1=-12.191＋0.972X,其中Y_0为预后良好,Y_1为预后不良。用该判别方程式对患者预后进行判别,正确率为71.8%,进一步行Bayes交互验证,计算准确度为71.7%,提示该判别方程式有较好的稳定性。结论 ASTRAL量表评分对后循环ACI患者预后有一定的评估价值。
Objective To investigate the value of ASTRAL scale in predicting the prognosis of posterior circulation acute cerebral infarction. Methods Enrolled 211 patients with posterior circulation acute cerebral infarction who were hospitalized in the Department of Neurology of Affiliated Hospital of North China University of Science and Technology from June 2011 to June2014. At 24 hours after admission,patients were given ASTRAL score and were given mRS score on 30 days after onset to predict the prognosis. According to mRS score, the patients were divided into favorable prognosis group and unfavorable prognosis group. ROC curves of ASTRAL scale predicting the prognosis of the patients were made,and AUC,the maximum Youden index and the ideal boundary value were calculated; with prognosis as dependent variable and ASTRAL score as independent variable,Fisher linear discriminant analysis was conducted. Results Favorable prognosis group had 167 patients, and unfavorable prognosis group had 44 patients. Favorable prognosis group was lower than unfavorable prognosis group in ASTRAL score（ t=- 7. 66,P〈0. 05）. With the increase of ASTRAL score, the rate of unfavorable prognosis increased significantly（ P〈0. 05）. AUC of ASTRAL predicting the prognosis of the patients was 0. 788,with 95% CI（ 0. 712,0. 864）,an ideal boundary value of 19,a sensitivity of 72. 7%,a specificity of 68. 9% and an accuracy of 69. 7%. According to ideal boundary value,the patients were divided into high- score group（ ≥19,n = 84） and low- score group（ 〈19,n = 127）. High- score group was higher than low- score group in the rate of unfavorable prognosis（ χ~2= 25. 14, P〈0. 01）. By Fisher linear discriminant analysis,discriminant equations were found as Y_0=- 6. 807 ＋ 0. 709 X and Y_1=- 12. 191 ＋ 0. 972 X,with Y_0 representing favorable prognosis and Y_1 representing unfavorable prognosis. The accuracy of the equation predicting patients＇ prognosis was71. 8% and the accuracy by further Bayes cross validation was 71. 7%,
出处
《中国全科医学》
CAS
CSCD
北大核心
2016年第9期1053-1055,共3页
Chinese General Practice
基金
河北省重大医学科研课题项目基金(zd2013089)
