摘要
目的评价中文版Barthel指数（BI）量表在我国卒中人群中的信度、效度和敏感度。方法将全国22家医疗中心符合纳入标准的537例中、重度缺血性卒中患者随机分为A组（274例）、B组（263例）,分别给予21d的不同治疗;由经过统一培训的研究人员于纳入研究第21、60、90天的3个时间点,采用中文版BI量表对卒中患者的日常生活活动能力进行测评。通过SPSS13.0统计软件包对有效的BI量表进行信度、效度和敏感度的评价,其中信度检验用Cronbach′sα系数、Guttman split-half和Spearson相关系数来评价。采用人为推理判断方法评价内容效度,采用因子分析法考核量表的结构效度;通过相关分析考察量表的效标效度。各领域敏感度通过配对t检验来评估。结果课题完成治疗共537例,在纳入研究第21、60、90天,分别回收的有效的BI量表为529（A组269份,B组260份）、526（A组269份,B组257份）、525（A组269份,B组256份）份。中文版BI量表内容一致信度的Cronbach′sα系数〉0.92,Guttman split-half系数在0.86-0.89之间,各领域的重测信度系数均〉0.82,显示了良好的信度。中文版BI量表与“准金标准”[（NIHSS）及改良Rankin量表（mRS）]呈适度的相关性（r=0.71-0.75,P〈0.001）,中文版BI量表各时间点的累计贡献率范围为75.03%-76.49%,显示了较好的效度。除A组第21与60天“可控制大便”（P=0.347）,B组“可控制大便”（P=0.095）、“可控制小便”（P=0.399）和B组第21与90天“可控制大便”（P=0.209）外,BI各项显示了良好的敏感度（P〈0.05）。结论中文版BI量表在中、重度卒中患者中,具有良好的信度、效度及敏感度,可作为卒中患者预后的评价指标。
Objective To evaluate the reliability, validity and sensibility of the Chinese Barthel Index （BI） in a Chinese stroke population.Methods A total of 537 patients with moderate to severe ischemic stroke, who were accorded with the inclusive criteria from 22 medical centers in China, were randomly allocated into groups A （n=274） and B （n=263）. They were treated with different therapeutic programmes for 21 days. The activity of daily living of patients with stroke were assessed by the unify-trained invesitgators with the Chinese BI at 3 time points （21, 30 and 90 days after inclusion）. The reliability, validity and sensibility of the effective Chinese BI were evaluated with SPSS 13.0, in which the reliability test was evaluated with Cronbach＇s α coefficient, Guttman split-half and Spearson correlation coefficient. The content validity was evaluated with the method of subjective reasoning and judging; the construction validity of the scales was assessed with factor analysis method; and the criterion validity of the scales was assessed with the correlation analysis. The sensitivity in all domains were assessed with the paired-samples T test.Results A total 537 patients were treated after the subject was completed. At day 21, 30 and 90 after inclusion, the effective BI scales were 529 （Group A 269 and Group B 260）, 526 （Group A 269 and Group B 257） and 525 （Group A 269 and Group B 256）, respectively. The Chinese BI showed favorable reliability in Cronbach＇s α coefficient of internal consistent reliability （〉0.92）, in Guttman split-half coefficient of internal consistent reliability （0.86 to 0.89）, and in retest reliability coefficient of every domain （〉0.82）. The Chinese BI were moderately correlated with ＂the quasi-gold standard＂ [（NIHSS and the modified Rankin scale （mRS） ] （r=0.71 to 0.75, P〈0.001）. The cumulative contribution rate of the Chinese BI at different time points ranged from 75.03% to 76.49%, which showed better validity. The Chinese BI showed ideal sensitivity in all domains at different time points （P〈0.05） except for ＂controllable bowel＂ （P=0.347） at day 21 and 60 in group A, ＂controllable bowel＂ （P=0.095） , ＂controllable bladder＂ （P=0.399） in group B, and ＂controllable bowel＂ （P=0.209） at day 21 and 90 in group B.Conclusion The Chinese BI has favorable reliability, validity and sensibility in patients with moderate and severe stroke, and it can be used as an evaluation index for prognosis in patients with ischemic stroke.
出处
《中国脑血管病杂志》
CAS
2007年第11期486-490,共5页
Chinese Journal of Cerebrovascular Diseases
基金
国家科技支撑计划重大项目(2006BAI04A02)
国家科技十五攻关项目(2004BA721A02
2001BA701A12b)
广东省科技厅科技攻关项目(2005B33001019)
广东省自然科学基金资助项目(06301402)
