Gianola Silvia, Guida Stefania, Ravot Gaia, Lunny Carole, Bargeri Silvia, Castellini Greta
IRCCS Istituto Ortopedico Galeazzi, Unit of Clinical Epidemiology, Milan, Italy.
University of Milan, Milan, Italy.
J Clin Epidemiol. 2025 Apr 9;183:111783. doi: 10.1016/j.jclinepi.2025.111783.
Network meta-analysis (NMA) is a method for comparing multiple interventions simultaneously, combining evidence to estimate and rank their relative effectiveness and safety across a network of studies. This study evaluates (i) epidemiological and descriptive characteristics, (ii) reporting completeness, and (iii) methodological quality of NMAs.
In this metaresearch study (protocol at https://osf.io/pa6dz/), we searched PubMed for systematic reviews with NMAs indexed in January 2023. We extracted epidemiological and descriptive data, assessed reporting completeness using the modified Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) extension for NMA, and evaluated the methodological quality using A Measurement Tool to Assess Systematic Reviews 2 (AMSTAR-2).
Among the 139 NMAs, 77% were published in specialty journals (median journal impact factor [JIF] 4), and 52% originated from China. Reporting completeness and methodological quality were generally of a medium quality, with the median NMAs fulfilling 71% of the modified Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Network Meta-Analyses (PRISMA-NMA) and 63% of the AMSTAR-2 criteria. Items such as "network geometry" for modified PRISMA-NMA (15%) and "list of excluded studies" for AMSTAR-2 (12%) were frequently unfulfilled. Better reporting and methodological quality were associated with registered protocol, non-Chinese country, higher JIF, and larger author teams.
We highlight gaps in both reporting and methodological quality in NMAs. We recommend future authors to plan and conduct NMAs within a large author team that includes statistical experts and to strictly adhere to reporting and methodological quality standards. More attention should be given to the reporting of network geometry and documenting the list of excluded studies.
网状Meta分析（NMA）是一种同时比较多种干预措施的方法，它整合证据以估计并排列它们在一系列研究中的相对有效性和安全性。本研究评估（i）流行病学和描述性特征，（ii）报告完整性，以及（iii）NMA的方法学质量。
在139项NMA中，77%发表于专业期刊（期刊影响因子[JIF]中位数为4），52%来自中国。报告完整性和方法学质量总体中等，NMA中位数满足修改后的网状Meta分析的系统评价和Meta分析首选报告项目（PRISMA-NMA）的71%以及AMSTAR-２标准的63%。修改后的PRISMA-NMA中的“网络几何结构”（15%）和AMSTAR-2中的“排除研究列表”（12%）等项目经常未得到满足。更好的报告和方法学质量与注册方案、非中国国家、更高的JIF以及更大的作者团队相关。
我们强调了NMA在报告和方法学质量方面的差距。我们建议未来的作者在一个包括统计专家的大型作者团队内计划和开展NMA，并严格遵守报告和方法学质量标准。应更多关注网络几何结构的报告以及记录排除研究的列表。
