护理研究中重复测量设计及其资料处理的考量

Repeated measurement design and data processing considerations in nursing research

[目的]描述我国护理研究中干预性重复测量设计的特点及统计学处理情况。[方法]以近6年公开发表的45篇干预性重复测量设计的论文为研究对象,采用内容分析法从研究设计特点与统计学处理情况两大方面进行分析。[结果]设计方案以随机对照设计为主(68.9%)。大多数(77.8%)后测在干预期间进行,而且后测非等距的情况占46.7%并无相应统计学处理。失访率的范围在2.6%~34.6%,失访情况占31.1%,主要处理方法为剔除失访资料。有2篇文献的重测资料涉及分类数据,且采用了χ2检验进行统计学处理。在存在组间-时间交互效应的37篇文献中,18篇未进行简单效应分析。[结论]随着我国护理研究的发展,护理工作者已经尝试采用干预性重复测量设计的方法,并能够正确选择重复测量资料方差分析进行处理,但在研究设计与统计学方面还存在诸多缺陷,科研水平有待提高。

Objective：To describe the characteristics and statistical treatment of interventional repeated measurement design in nursing research in China.Methods：A total of 45 paper on interventional repeated measurement design published in the last six years were involved as the research object.Content analysis method was used to analyze 2 aspects：the characteristics of research design and statistical processing.Results：The design scheme was dominated by randomized controlled design（68.9%）.The majority（77.8%）of the posttest was conducted during the intervention period,and the non-isometric post-test condition accounted for 46.7% with no corresponding statistical treatment.The loss rate ranged from 2.6%to 34.6%,and the loss of visit accounted for 31.1%,with eliminating the missing information as the main processing method.The retest data of 2 pieces of literatures involved classification data,and the chi-square test was used for statistical processing.In the 37 literatures with intergroup-time interaction effect,18 paper did not have simple effect analysis.Conclusions：With the development of nursing research in China,nursing researchers tried using interventional repeated measurement design method,and were able to choose correct repeated measurement data analysis of variance for processing.While there were many defects in the study design and statistics,the research level needs to be improved.