| 中文关键词: 乳腺肿瘤 筛查 系统评价再评价 质量评估 证据分级 |
| 英文关键词:breast neoplasms screening overview of review quality assessment evidence classified |
| 基金项目:北京市优秀人才培养资助青年拔尖团队项目（20170000212 23TD05）；国家自然科学委员会项目（81602930） |
|
| 摘要点击次数: 2241 |
| 全文下载次数: 422 |
| 中文摘要: |
| 摘 要：［目的］ 评价乳腺癌筛查系统评价的方法学质量及其结论的可靠程度。［方法］ 计算机检索PubMed、EMBASE、The Cochrane Library（2018年第2期）、Web of Knowledge、WanFang、CNKI 和CBM数据库，纳入乳腺癌筛查相关的系统评价，检索时限截止为2018年2月28日。采用ROBIS工具评价纳入研究的方法学质量，并基于GRADE方法对系统评价纳入的结局指标进行证据质量分级。［结果］ 共纳入13个系统评价。ROBIS工具评价结果显示，阶段一“目标问题与系统评价问题”的四个问题，有8个研究是完全符合。阶段二，系统评价制作过程中偏倚风险程度的评价结果为：研究的纳入排除标准领域中，8个研究偏倚风险程度为“低”，2个为“高”，3个“不确定”；研究的检索和筛选领域中，7个研究偏倚风险程度为“低”，4个研究为“高”，2个“不确定”；数据提取和质量评价领域中，6个研究偏倚风险程度为“低”，7个研究为“高”；数据合成和结果呈现领域中，6个研究偏倚风险程度为“低”，6个研究偏倚风险程度“高”，1个“不确定”。阶段三，偏倚风险的评价结果为：7个系统评价为“低”风险偏倚，6个系统评价为“高”风险偏倚。运用GRADE分级结果显示，乳腺癌筛查降低死亡率结局指标中2个研究的证据质量为A级，2个为B级，3个由于不精确性和发表偏倚而为C级（低质量）证据。乳腺癌检出率中1个研究为A级证据，1个为B级，其余2个研究为C级。假阳性的2个研究的证据质量为C级。过度筛查中2个系统评价均为C级。筛查潜在危害中1个研究的证据级别为B级。［结论］ 当前针对乳腺癌筛查系统评价的方法学质量总体尚可，但结论的证据水平较低，对筛查终点结局的数据收集不足。因此，临床医生在使用这些证据进行临床决策时尚需依据当地实际情况讨论后应用。 |
| 英文摘要: |
| Abstract：［Purpose］ To evaluate the quality of methodology and evidence in systematic reviews(SRs) of breast cancer screening. ［Methods］ Electronic searches were carried out in the PubMed，The Cochrane Library(Issue 2，2018)，EMBASE，Web of Knowledge，CBM，WanFang Data and CNKI to collect SRs of breast cancer screening from inception to 28 February 2018. The ROBIS tool was applied for grading quality of methodology and the GRADE system was applied for grading quality of evidence of included SRs. ［Results］ A total of 13 SRs including 5 outcomes(mortality，detection rate，false-positive，over-diagnosis and potential benefits and harms) were included. The ROBIS evaluation showed that 8 studies were completely matched the 4 questions of phase Ⅰ；in the phase Ⅱ，8 studies were in low risk of bias in the including criteria field，7 studies were low risk of bias in the literature search and screening field，6 studies were in low risk of bias in the data abstraction and quality assessment field，and 6 studies were in low risk of bias in the data synthesis field；in the phase Ⅲ，7 studies were in low risk. The GRADE classification showed that 2 studies had level A evidence on the outcome of mortality；1 study had level A evidence on detection；others had level B and C on evidence. ［Conclusion］ The quality of methodology of SRs is modest；while the quality of evidence is low. The outcomes of screening have not been collected sufficiently. Therefore，when clinicians make clinical decisions they have to use these evidence with the local context. |
|
在线阅读
查看全文 查看/发表评论 下载PDF阅读器 |
