中文发表的乳腺癌随机对照试验摘要报告质量调查

目的 评价中文发表的乳腺癌随机对照试验(RCT)摘要报告质量。 方法 运用计算机检索主要中文数据库，纳入中文发表的乳腺癌RCT，采用“摘要CONSORT”标准评价纳入RCT摘要的报告质量，使用Meta-Analyst软件进行统计分析。 结果 最终纳入符合标准的RCT 965篇，评价结果显示:乳腺癌RCT摘要中存在严重报告质量缺陷，尤其体现在“标题中判定为随机研究”(1.1%,95%CI:0.6%～2.0%)、“随机化”(3.8%,95%CI:2.8%～5.2%)、“盲法”(1.0%,95%CI:0.6%～1.9%)、“统计方法”(11.5%,95%CI:9.6%～13.7%)、“随机分组的例数”(7.5%,95%CI:6.0%～9.3%)、“招募对象”(12.2%,95%CI:10.3%～14.4%)、“试验注册信息”(13.1%,95%CI:11.1%～15.3%)和“资助来源”(8.1%,95%CI:6.5%～10.0%)8个方面。“摘要CONSORT”发布以后，可以显著提高“研究对象”“干预”“统计方法”“结局”和“结论”的报告率。比较非中国科学引文数据库(CSCD)，CSCD收录RCT的摘要在“标题识别随机”“随机化”“招募对象”和“资助来源”的报告上差异有统计学意义。 结论 中文发表乳腺癌RCT摘要存在严重报告缺陷，应当引起期刊编辑和研究员的重视。 

Evaluation of quality of randomized controlled trials research papers among Chinese abstracts for breast cancer

Objective To assess the quality of randomized controlled trials research papers among Chinese abstracts for breast cancer. Methods A systematical databases search was conductedto identify papers about RCT of breast cancer published in China."Abstract-CONSORT" statement was used to assess the quality of reporting.Statistical analysis was performed using Meta-Analyst software. Results A total of 965 RCT were identified.The serious reporting flaws were appeared for RCT abstract of breast cancer,especially,regarding following items:thestudy identified as randomized in title(1.1%,95%CI:0.6%-2.0%),randomization 3.8%,95%CI:2.8%-5.2%),blinding(1.0%,95%CI:0.6%-1.9%),statistical methods(11.5%,95%CI:9.6%-13.7%),numbers randomized(7.5%,95%CI:6.0%-9.3%),participants (12.2%,95%CI:10.3%-14.4%),trial registration(13.1%,95%CI:11.1%-15.3%),funding (8.1%,95%CI:6.5%-10.0%).The improvements in participants,intervention,statisticalmethods,results and conclusions were increased when abstract CONSORT was published.There was statistical significant difference in study identified as randomized in title,randomization,and participants and funding betweenRCT abstract recorded by CSCD andnon-CSCD (international papers). Conclusion The serious reporting flaws were appeared for RCT abstract of breast cancer published in Chinese.More attentions should be made for editors and authors.