论文部分内容阅读
在医学研究中,除了计量资料、计数资料外,还有一类等级分组资料。例如化验结果、临床疗效分析的等级分组。参照单位(Ridit)分析法是一种关于等级资料进行对比组与标准组比较的假设检验方法。其基本思想是:先确定一个标准组(常为以往积累的资料或样本含量相当大的资料)作为特定总体,求得各等级的R值。标准组R值的均数R标为0.5。假设对比组是来自该特定总体的随机样本,则可信度为(1—α)时,对比组总体R值的可信区间包括0.5的概率为(1—α);反之,若此区间不包括0.5,则可根据检验水准α拒绝假设,而认为对比总体与标准总体有差别。本文利用Ridit分析法评价医院疗效质量,借以说明Ridit分析法在医院质量管理中的实用价值。
In medical research, apart from measurement data and count data, there is also a class of hierarchical group data. For example, the grouping of test results and clinical efficacy analysis. The reference unit (Ridit) analysis is a hypothesis testing method that compares rank data with the standard group. The basic idea is to first determine a standard group (usually data accumulated in the past or data with a large sample size) as a specific population and obtain the R values for each level. The mean R of standard group R values is 0.5. Assuming that the comparison group is a random sample from this particular population, when the confidence level is (1-α), the probability that the confidence interval of the overall group R value of the comparison group includes 0.5 is (1-α); conversely, if this interval is not Including 0.5, you can reject the hypothesis based on the inspection level α, and consider that there is a difference between the comparison population and the standard population. This article uses Ridit analysis to evaluate the efficacy of hospitals in order to illustrate the practical value of Ridit analysis in hospital quality management.