【摘要】 本文将归因比例的概念推广到结果变量为有序多分类的情况，基于超额危险的概念和有序结果变量的特性定义了类别归因比例的概念。旨在定量反映归因于某暴露因素的特定结果分类的病例比例。同时给出了类别归因比例的点估计和方差估计，并应用实例说明了新概念的重要性。更多还原

【Abstract】 Objective: Objectie: Categorical Attributable Fraction is presented to quantify the the proportion of diseased individualreduced in certain response category when certain exposure is removed. Methods: The concept of categorical attributable fraction (CAJj) is defined on excess risk and characteristic of ordinal data. Delta method was used toderived asymptotic variance of the measure. Results in the evaluation of the effect of midline episiotomy on degree oflaceration, four Categorical Attributable Fraction and asyrnptotic variance was calculated. conclusion: Categorical Attributable Fraction (CAFj) can be used to assess the public health consequences of an association between an exposurefactior and certain category of disease.更多还原