Ottawa Methods Centre, Clinical Epidemiology Program, Centre for Practice-Changing Research, Ottawa Hospital Research Institute, Ottawa, Ontario, Canada.
Translational Research in Biomedicine (TRIBE) Program, University of Split School of Medicine, Split, Croatia.
Syst Rev. 2017 Jun 19;6(1):117. doi: 10.1186/s13643-017-0507-6.
The methodological quality and completeness of reporting of the systematic reviews (SRs) is fundamental to optimal implementation of evidence-based health care and the reduction of research waste. Methods exist to appraise SRs yet little is known about how they are used in SRs or where there are potential gaps in research best-practice guidance materials. The aims of this study are to identify reports assessing the methodological quality (MQ) and/or reporting quality (RQ) of a cohort of SRs and to assess their number, general characteristics, and approaches to 'quality' assessment over time.
The Cochrane Library, MEDLINE®, and EMBASE® were searched from January 1990 to October 16, 2014, for reports assessing MQ and/or RQ of SRs. Title, abstract, and full-text screening of all reports were conducted independently by two reviewers. Reports assessing the MQ and/or RQ of a cohort of ten or more SRs of interventions were included. All results are reported as frequencies and percentages of reports.
Of 20,765 unique records retrieved, 1189 of them were reviewed for full-text review, of which 76 reports were included. Eight previously published approaches to assessing MQ or reporting guidelines used as proxy to assess RQ were used in 80% (61/76) of identified reports. These included two reporting guidelines (PRISMA and QUOROM) and five quality assessment tools (AMSTAR, R-AMSTAR, OQAQ, Mulrow, Sacks) and GRADE criteria. The remaining 24% (18/76) of reports developed their own criteria. PRISMA, OQAQ, and AMSTAR were the most commonly used published tools to assess MQ or RQ. In conjunction with other approaches, published tools were used in 29% (22/76) of reports, with 36% (8/22) assessing adherence to both PRISMA and AMSTAR criteria and 26% (6/22) using QUOROM and OQAQ.
The methods used to assess quality of SRs are diverse, and none has become universally accepted. The most commonly used quality assessment tools are AMSTAR, OQAQ, and PRISMA. As new tools and guidelines are developed to improve both the MQ and RQ of SRs, authors of methodological studies are encouraged to put thoughtful consideration into the use of appropriate tools to assess quality and reporting.
系统评价(SRs)的方法学质量和报告完整性是实施循证医疗保健和减少研究浪费的基础。虽然存在评估 SRs 的方法,但对于这些方法在 SRs 中的使用情况或研究最佳实践指导材料中潜在的差距知之甚少。本研究的目的是确定评估一组 SRs 的方法学质量(MQ)和/或报告质量(RQ)的报告,并评估随着时间的推移这些报告的数量、一般特征和“质量”评估方法。
从 1990 年 1 月至 2014 年 10 月 16 日,对 Cochrane 图书馆、MEDLINE®和 EMBASE®进行了检索,以查找评估 SRs 的 MQ 和/或 RQ 的报告。两名审查员分别对所有报告的标题、摘要和全文进行了筛选。纳入了评估十项或更多干预措施的 SRs 的 MQ 和/或 RQ 的报告。所有结果均以报告的频率和百分比报告。
从 20765 个唯一记录中检索到 1189 个进行全文审查,其中 76 个报告被纳入。在确定的报告中,有 80%(61/76)使用了以前发表的评估 MQ 或报告指南的八种方法来评估 RQ。这些方法包括两个报告指南(PRISMA 和 QUOROM)和五个质量评估工具(AMSTAR、R-AMSTAR、OQAQ、Mulrow、Sacks)和 GRADE 标准。其余 24%(18/76)的报告则制定了自己的标准。PRISMA、OQAQ 和 AMSTAR 是最常用于评估 MQ 或 RQ 的已发表工具。在与其他方法结合使用的情况下,有 29%(22/76)的报告使用了已发表的工具,其中 36%(22/76)评估了对 PRISMA 和 AMSTAR 标准的依从性,26%(6/22)使用了 QUOROM 和 OQAQ。
评估 SRs 质量的方法多种多样,没有一种方法被普遍接受。最常用的质量评估工具是 AMSTAR、OQAQ 和 PRISMA。随着新的工具和指南的制定以提高 SRs 的 MQ 和 RQ,鼓励方法学研究的作者在使用适当的工具评估质量和报告时进行深思熟虑。