配对相关数据的统计学方法

Statistical methods for the bilateral correlated data

配对相关数据经常在医学研究中应用,例如,在眼科或耳鼻喉科研究中,对配对器官中每个样本信息进行数据分析.配对器官的测量数据通常具有高度相关性,而大多数统计推断方法假定样本的观测值是相互独立的.研究表明,忽略配对数据的组内相关性会导致显著性水平的增加.有很多统计方法来解决配对数据的组内相关性.此外,忽略配对数据的相关性或混杂效应可导致结果发生偏差,因此,调整和控制统计推断中的混杂效应至关重要.本文回顾讨论这些统计方法,并提出统计推断方法的建议.

Bilateral correlated data are often encountered in medical researches such as ophthalmologic(or otolaryngologic) studies, in which each unit contributes information from paired organs to the data analysis.Measurements obtained from paired organs of a subject are generally highly correlated, whereas many statistical tests assume observations in a sample are independent. Previous studies showed that the statistical inference for bilateral correlated data ignoring the presence of intra-class correlation could lead to inflated significance levels.Various statistical methods have been developed to tackle this intra-class correlation on the bilateral correlated data analysis. Furthermore, in some studies, it is very important to adjust the effect of confounder on statistical inferences, since either ignoring the intra-class correlation or confounding effect may lead to biased results. In this paper, we review these methods, discuss the applications and provide statistical advice.