随机对照试验偏倚风险评估方法及其运用情况调查  

Investigation of Tools for Assessing Risk of Bias in Randomized Controlled Trials and Their Applications

作　　者：许卫华 XU Wei-Hua(Center for Clinical Research Design,Measurement and Evaluation,Institute of Clinical Pharmacology,Guangzhou University of Chinese Medicine,Guangzhou 510405 Guangdong,China)

机构地区：[1]广州中医药大学临床药理研究所DME中心

出　　处：《广州中医药大学学报》2019年第9期1452-1458,共7页Journal of Guangzhou University of Traditional Chinese Medicine

基　　金：广东省研究生示范课程建设项目（编号：2016SFKC-10）

摘　　要：【目的】了解随机对照试验偏倚风险评估方法及运用情况。【方法】通过检索中国生物医学期刊数据库2017年1月至2018年6月收录的干预措施随机对照试验系统评价和Meta分析文献,选取发表于《中国循证医学杂志》的所有文献以及发表于其他中文核心期刊的部分文献为调查样本,收集调查样本中偏倚风险评估采用的方法、评估的内容、判断依据,以及评估结果的报告及应用等信息。【结果】《中国循证医学杂志》和其他核心期刊分别有66篇和109篇文献(共计175篇)纳入分析,文献类型均为Meta分析。所有发表于《中国循证医学杂志》的Meta分析均采用Cochrane协作网偏倚风险评估工具(RoB);发表于其他核心期刊的文献中有74篇(67.9%)采用该工具,25篇(22.9%)采用Jadad量表,2篇(1.8%)同时采用上述两种方法,2篇(1.8%)用其他方法,5篇(4.6%)未说明评估方法,1篇(0.9%)未进行偏倚风险评估。在使用2011年版RoB的Meta分析中分别有54.3%(25/46)的《中国循证医学杂志》文献和30.4%(14/46)的其他中文核心期刊文献未根据施盲对象的不同分开评估"盲法"相关偏倚;超过92%的Meta分析评估了"报告偏倚"及"其他偏倚";发表于其他中文核心期刊的Meta分析中有4篇基于结局指标分别评估"测量偏倚"、"实施偏倚",各有1篇Meta分析将偏倚风险评估结果作为纳入标准及敏感性分析依据。【结论】RoB为随机对照试验偏倚风险评估的最常用方法,但实际运用时与该工具的操作指引存在偏离。Objective To investigate the tools for assessing the risk of bias in randomized controlled trials and their current application in systematic reviews published in Chinese journals.Methods We reviewed all of the systematic reviews published in Chinese Journal of Evidence-Based Medicine(CJEBM),and randomly selected systematic review from other Chinese Core Journals(CCJ)recruited in the Chinese Biomedical Literature Database from January,2017 to June,2018.Information was extracted from articles,including the tools for assessing the risk of bias,components contained in the tools,criteria for the risk judgment,presentation and utilization of the assessment results of risk of bias.Results A total of 175 systematic reviews were included,66 published in CJEBM and 109 from CCJ.All of the CJEBM’s reviews adopted the Cochrane Collaboration’s tool for assessing risk of bias in randomized clinical trials(RoB)as the risk assessment tool,while 74 reviews from CCJ(accounting for67.9%)adopted RoB.For the rest reviews from CCJ,25 reviews(22.9%)used modified Jadad scale,2(1.8%)used both RoB and modified Jadad scale,2(1.8%)adopted other methods,5(4.6%)had no description of the assessment tools,and one review(0.9%)did not report the risk of bias.Among the reviews using the 2011 version RoB,54.3%(25/46)CJEBM’s reviews and 30.4%(14/46)CCJ’s reviews merged the blinding of patients and care providers with the blinding of outcome assessors;over 92%of the reviews presented"reporting bias"and"other bias"domain;4 reviews from CCJ had the report of"detection bias"and"performance bias",and there were 2 Meta-analysis articles using the risk of bias assessment as a basis for sensitivity analysis or as an exclusion criterion.Conclusion Although the Cochrane tool RoB is the most common approach to assess the risk of bias in the randomized controlled trials,its application still frequently deviates from the guideline of RoB.

关 键 词：偏倚风险 评估方法 Cochrane协作网随机对照试验偏倚风险评估工具 系统评价 META分析 

分 类 号：R2-03[医药卫生—中医学]