论文部分内容阅读
目的:进一步筛选基于冠心病心绞痛患者报告的临床结局评价量表的条目。方法:多中心收集了319例经冠脉造影证实的患者,在此资料的基础上采用条目分布考察法、离散趋势法、因子分析法、克朗巴赫系数法、逐步回归法、逐步判别法6种方法分别从不同角度来筛选量表的条目。结果:采用条目分布考察法从集中趋势的角度筛选出33个条目,采用离散趋势法从敏感性角度筛选34个条目,采用因子分析法从条目的代表性角度筛选26个条目,采用克朗巴赫系数法从内部一致性的角度筛选出27个条目,采用逐步回归法筛选出了影响性较大的4个条目,采用判别分析法从体现区分度的角度筛选出6个条目。结论:患者报告的临床结局评价量表的条目筛选可以在多中心临床流行病学资料的基础上结合多种统计方法较科学的进行。
OBJECTIVE: To further screen entries based on clinical outcome measures reported by patients with coronary heart disease. Methods: A total of 319 patients confirmed by coronary angiography were collected from multiple centers. Based on this data, six methods were used: entry distribution investigation, discrete trend analysis, factor analysis, Cronbach’s coefficient, stepwise regression and stepwise discrimination Methods Screen the entries of the scale from different perspectives. Results: 33 items were screened from the perspective of concentration trend by using item distribution survey method, 34 items were screened from the perspective of sensitivity using discrete trend method, 26 items were screened from the representative point of item using factor analysis method, Cronbach’s coefficient The law screened 27 items from the perspective of internal consistency, screened out four items with more influential by stepwise regression, and used discriminant analysis to select 6 items from the perspective of distinguishing degree. CONCLUSIONS: Screening of entries in the clinical outcome rating scale reported by patients may be scientifically based on multicenter clinical epidemiological data combined with various statistical methods.