Clinic for Palliative Care, Medical Center, University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany.
Institute of Medical Biometry and Statistics, Faculty of Medicine and Medical Center, University of Freiburg, Freiburg, Germany.
J Clin Epidemiol. 2021 Aug;136:84-95. doi: 10.1016/j.jclinepi.2021.03.010. Epub 2021 Mar 16.
To assess the methodological quality and the consideration of heterogeneity in systematic reviews (SRs).
We conducted a methodological study (CRD42019134904) and searched three databases from January 2010 to July 2019. Interventional SRs with a statistically significant meta-analysis of at least four randomized controlled trials in advanced cancer patients were included. A MeaSurement Tool to Assess Systematic Reviews (AMSTAR) 2 was used to evaluate the SRs' methodological quality. The consideration of heterogeneity was categorized in clinical or/and methodological heterogeneity and not explored.
From 6234 identified references, 261 SRs were included. Most SRs had a critically low quality (230, 88.1%). The majority of them (209, 80.1%) was classified as critically low because of non-registration (222, 85.1%) combined with the non-reporting of excluded full-texts and missing justifications for exclusion (218, 83.5%). Heterogeneity in trial results was not explored at all in 51 (19.5%) SRs whereas clinical heterogeneity was considered in 117 (44.8%), methodological heterogeneity in 13 (5.0%), and both clinical and methodological heterogeneity in 80 (30.7%) SRs.
The consideration of these findings in trainings for review authors and peer reviewers could improve the awareness of quality criteria and the quality of future SRs.
PROSPERO-ID: CRD42019134904.
评估系统评价（SR）的方法学质量和对异质性的考虑。
我们进行了一项方法学研究（CRD42019134904），并于 2010 年 1 月至 2019 年 7 月检索了三个数据库。纳入了对晚期癌症患者至少有 4 项随机对照试验进行统计学显著荟萃分析的干预性 SR。使用评估系统评价的测量工具（AMSTAR 2）来评估 SR 的方法学质量。对异质性的考虑分为临床或/和方法学异质性，未进行探索。
从 6234 篇鉴定文献中，纳入了 261 篇 SR。大多数 SR 的质量为临界低（230 篇，88.1%）。其中大多数（209 篇，80.1%）被归类为临界低，原因是非注册（222 篇，85.1%）与未报告排除的全文和缺失排除的理由相结合（218 篇，83.5%）。51 篇（19.5%）SR 根本没有探讨试验结果的异质性，117 篇（44.8%）考虑了临床异质性，13 篇（5.0%）考虑了方法学异质性，80 篇（30.7%）考虑了临床和方法学异质性。
在对综述作者和同行评审者的培训中考虑这些发现，可以提高对质量标准和未来 SR 质量的认识。
PROSPERO-ID：CRD42019134904。
