动物实验系统评价/Meta分析的质量和报告特征

Methodological and reporting quality of systematic review/meta-analysis of animal studies

目的对动物实验系统评价/Meta分析(SRs/MAs)的研究现状和质量进行分析。方法计算机检索Pub Med、EMbase、The Cochrane Library、BIOSIS Previews、VIP、CNKI、CBM和Wan Fang Data数据库,搜集国内外发表的动物实验SRs/MAs,检索时限均为建库至2016年6月。由2名研究者独立筛选文献、提取资料,并对纳入研究进行描述性分析。结果最终纳入609篇动物实验SRs/MAs,分别来自27个国家、发表在526种期刊。在方法学质量方面,仅36.8%(224/609)的研究对纳入原始动物实验的偏倚风险进行了评估。在报告质量方面,文献筛选方法(41.9%,255/609)、资料提取方法(32.0%,195/609)、提供纳入研究的基本特征(41.2%,251/609)的报告率均低于50%。结论目前已发表动物实验SRs/MAs在方法和报告方面均存在一定的问题。有必要通过开展动物实验SRs/MAs方法学培训,促进其制作流程的规范化和科学性。同时采用科学的方法制定动物实验SRs/MAs的报告规范,并制定相关政策,促进其在相关专业期刊稿约中的引入,提高报告质量。

Objectives To survey the current research situation, methodological and reporting quality of the systematic review/meta-analysis （SRs/MAs） of animal studies. Methods PubMed, EMbase, BIOSIS Previews, CNKI, WanFang Data, CBM and VIP databases were searched to collect SRs/MAs of animal studies from inception to June 2016. Two reviewers independently screened literature, extracted data, a descriptive analysis was then conducted. Results A total of 609 SRs/MAs of animal studies were included, which were from 27 countries and published in 526 journals.Merely 36.8% （224/609） studies assessed the risk of bias in the original animal experiments. Less than 50% studies reported the method of literature selection （41.9%, 255/609）, data abstraction （32.0%, 195/609） and study characteristics （41.2%, 251/609）. Conclusions The published SRs/MAs of animal studies is poor in both methodological and reporting quality. Thus, we hope to improve awareness and actual use rates of these guidelines by basic medical researchers and journal editors, thereby improving the quality of animal experimental methods and reporting standards.