脑卒中患者临床神经功能缺损程度评分的信度与效度

目的 评价脑卒中患者临床神经功能缺损程度评分的信度和效度.方法 222例急性脑卒中患者,分别由不同评定者于入院当天、入院第14天、发病后第90天对同一例脑卒中患者进行临床神经功能缺损程度评分.重测信度与评定者间信度用Kappa值表示,分半信度、内部一致性信度用Croubach α值表示.同时效度、预测效度采用Spearman相关分析.结构效度采用因子分析的方法,其中脑梗死患者根据英国牛津郡社区脑卒中项目(OCSP)分型评定3种不同梗死类型的临床神经功能缺损程度评分的结构效度.结果 所有条目重测信度均>0.6,评定者间信度中条目"步行能力"为0.542,分半信度0.911,Cronbach α值为0.886,不同评定方式间信度显示"上肢肌力"、"手肌力"的信度分别为0.393、0.386.其与NIHSS评定结果呈正相关(均P=0.000),不同OCSP分型患者神经功能缺损程度评分与NIHSS评分结果均相关,且与发病后90 d的Barthel指数和改良Rankin评分均相关.结论 临床神经功能缺损程度评分内部一致性信度较高,其与NIHSS有同时效度,对脑卒中预后有预测效度;对部分前循环梗死患者的结构效度最好,可以用来评定椎-基底动脉系统的脑卒中,但并不敏感.

Study on reliability and validity of the clinical neurologic deficit scale in patients with stroke

Objective To study the reliability and validity of the clinical neurologic deficit scale in evaluating stroke patients. Methods A total of 222 inpatients with acute stroke onset were included in the study. They were assessed when admitted, at the 14th and 90th day of hospitalization by different physicians using the clinical neurologic deficit scale. Intrarater and interrater reliability were determined using Kappa correlation. The split-half rehability and internal consistency were evaluated using Cranbach's a coefficient. Concurrent validity and the predictive validity were determined by spearman rank correlation coefficients. Construct validity was assessed by the factor analysis and the construct validity of the scale was measured according to the classifications of the Oxfordshire Community Stroke Project ischemic stroke subtypes in the patients with cerebral infarction. Results The scores of intrarater reliability in all items were higher than 0.6, the score of interrater reliability in the item "walking" was 0.542, the split-half reliability and the internal consistency were good as demonstrated by the score of 0.911 and 0.886 respectively, and assessment of reliability of different methods showed that "strength in upper limb" and "strength in hand", were poor as shown by the score of 0.393 and 0.386 respectively. The scale is highly correlated with the NIHSS ( both P=0.000) in both total and subtypes of stroke patients according to the classifications of the Oxfordshire Community Stroke Project by concurrent validity analysis. There was a high correlation between the scores of the scale and Barthel Index and the modified Rankin scales at the 90th day of hospitalization (both P=0.000). Conclusions The clinical neurologic deficit scale has a good internal consistency. There is concurrent validity between the scale and the NIHSS and could predict stroke outcome. Factor analysis of the scale displays the best construct validity in the patients with partial anterior circulation infarction, and could be used to evaluate the focus of vertebrobasilar artery despite its insensitivity.