Yu Nanze, Yu Panxi, Long Xiao, Huang Jiuzuo, Jia Yihong, Wang Xiaojun
From the *Division of Plastic Surgery, Peking Union Medical College Hospital, and †Plastic Surgery Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China.
Ann Plast Surg. 2017 Jan;78(1):111-118. doi: 10.1097/SAP.0000000000000757.
Meta-analyses are considered to be an important source of evidence. This review aims to systematically assess the quality of meta-analyses addressing topics in plastic surgery.
Electronic databases were selected for systematic review. A search was performed focusing on communication addresses containing terms related to plastic surgery, and detailed inclusion criteria were used. Related data were extracted and recorded according to the items of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. To assess the quality of the meta-analyses over time, studies published before and after PRISMA were evaluated.
A total of 116 meta-analyses were included. There was 1 study that was fully in compliance with the PRISMA items. The main flaws impacting the overall quality of the included studies were in the following areas: structured summary (48%), protocol and registration (2%), full electronic search strategy (35%), risk of bias in individual studies (41%), additional analyses (27%), risk of bias within studies (47%), additional analysis (30%), and funding (47%). Study quality was evaluated using relative risks (RR) with a 95% confidence interval (95% CI); this revealed that there were few significant improvements in adherence to the PRISMA statement after its release, especially in selection (RR, 1.80; 95% CI, 1.08-2.99), results of individual studies (RR, 2.88; 95% CI, 1.41-5.91), synthesis of results (RR, 3.08; 95% CI, 1.32-7.17), and funding (RR, 1.65; 95% CI, 1.21-2.24).
There have been measurable improvements in the quality of meta-analyses over recent years. However, several serious deficiencies remain according to the PRISMA statement. Future reviewers should pay more attention to not only reporting the main findings but also encouraging compliance with proper standards.
荟萃分析被视为重要的证据来源。本综述旨在系统评估整形外科学术主题的荟萃分析质量。
选择电子数据库进行系统综述。围绕包含整形手术相关术语的通讯地址展开搜索，并采用详细的纳入标准。根据系统评价和荟萃分析的首选报告项目（PRISMA）声明的项目提取并记录相关数据。为评估不同时期荟萃分析的质量，对PRISMA发布前后发表的研究进行了评估。
共纳入116项荟萃分析。仅有1项研究完全符合PRISMA项目要求。影响纳入研究整体质量的主要缺陷体现在以下方面：结构化摘要（48%）、方案与注册（2%）、完整电子检索策略（35%）、单个研究的偏倚风险（41%）、额外分析（27%）、研究内偏倚风险（47%）、额外分析（30%）以及资金资助（47%）。使用相对风险（RR）及95%置信区间（95%CI）对研究质量进行评估；结果显示，PRISMA声明发布后，在遵循该声明方面几乎没有显著改善，尤其是在选择（RR，1.80；95%CI，1.08 - 2.99）、单个研究结果（RR，2.88；95%CI，1.41 - 5.91）、结果合成（RR，3.08；95%CI，1.32 - 7.17）以及资金资助（RR，1.65；95%CI，1.21 - 2.24）方面。
近年来荟萃分析的质量有了可衡量的提高。然而，根据PRISMA声明，仍存在一些严重不足。未来的综述者不仅应更多关注报告主要研究结果，还应鼓励遵循适当的标准。
