1998-2014年《中华医学杂志》发表的Meta分析文献质量评价  

A systematic evaluation on the quality of Meta-analysis in articles published in National Medical Journal of China from 1998 to 2014

作　　者：韦所苏[1] 韦挥德[1] 黄晓红[1] 谭毅[1] 杨光和[1] 刘京虹[1] 吕文娟[1] 韦颖[1] 蓝斯琪[1] 刘慧[1] 吴腾燕[2] 

机构地区：[1]广西壮族自治区人民医院《中国临床新医学》杂志编辑部,南宁530021 [2]广西壮族自治区疾病预防控制中心结核病防制所

出　　处：《中华医学杂志》2015年第8期621-626,共6页National Medical Journal of China

摘　　要：目的 评价《中华医学杂志》发表的Meta分析文献的方法学质量和报告质量.方法 计算机检索及手工检索1998年1月至2014年10月《中华医学杂志》公开发表的Meta分析文献,采用Meta分析方法学质量评价工具（AMSTAR）量表进行方法学质量评价,采用Meta分析优先报告条目（PRISMA）量表进行报告质量评价.结果 共纳入文献74篇.纳入文献的方法学质量评分为3～10（中位数7）分,高质量者10篇（13.5％）,中等质量者61篇（82.4％）,低质量者3篇（4.1％）,各年AMSTAR得分比较差异无统计学意义（x2=10.205,P=0.423）;其中“是否提供了纳入和排除的研究文献清单”、“是否说明相关利益冲突”、“发表情况是否已考虑在纳入标准中,如灰色文献”及“是否评估了发表偏倚的可能性”符合率较低.报告质量评分为5.5～26（中位数20）分,≤15分有9篇（12.2％）,15.5 ～21分有39篇（52.7％）,21.5 ～27分有26篇（35.1％）,各年PRISMA得分比较差异有统计学意义（F=4.301,P=0.000）,2010年得分最高.其中结构式摘要、引言目的,方法中的方案和注册、资料提取、资料条目、单个研究存在的偏倚、研究间偏倚、补充分析,结果中的研究筛选、研究内部偏倚风险、研究间偏倚、补充分析,讨论中的研究局限性、结论、资金支持等报告不够全面.结论 《中华医学杂志》发表的Meta分析文献方法学质量和报告质量总体较高,可为临床决策提供良好的证据,但仍需按照Meta分析的写作要求不断提高.Objective To assess qualities on methodologies and reporting on Meta-analysis used in papers being published in National Medical Journal of China.Methods Computerized literature searching was performed in Wangfang Medical Online to collect articles that Meta-analysis was used in the National Medical Journal of China since January 1998 until October 2014.Manual retrieval was also conducted.Qualities on methodologies and reporting were evaluated by both Assessment of Multiple Systematic Reviews （AMSTAR） and Preferred Reporting Items for Systematic Reviews and Meta-Analyses （PRISMA） scales.Results A total of 74 papers were included.The results on the qualities of methodology evaluation in these papers were 3 to 10 （median 7） scores,10 articles （13.5%） were rated as high,61 articles （82.4%） as moderate and 3 articles （4.1%） as low.No statistically significant difference existed in each year of AMSTAR score （χ^2 =10.205,P =0.423）.The titles of AMSTAR scales with a lower coincidence rate were “Was a list of studies （included and excluded） provided？”,“Was the conflict of interest stated？”,“Was the status of publication （i.e.grey literature） used as an inclusion criterion？” and “Was the likelihood of publication bias assessed？”.The results on the qualities of reporting evaluation in these papers were 5.5 to 26 （median 20） scores,9 articles （12.2%） scored 〈 15 points,39 articles （52.7%） 15.5-21 points and 26 articles （35.1%） 21.5-27 points.There was statistically significant difference in each year of PRISMA score （F =4.301,P =0.000）.And the year 2010 was the highest one.The titles of PRISMA scales with a lack of comprehensive reports were “ Structured summary”,“Objectives of introduction”,“Protocol ＆ registration”,“Data collection process”,“Data items”,“Risk of bias in individual studies”,“Risk of bias across studies”,“Additional analyses”,“Study selection”,“Risk of bias within studies”,“Conclusi

关 键 词：META分析 质量评价 中华医学杂志 

分 类 号：G353.1[文化科学—情报学] R-03[医药卫生]