Bristol Medical School, University of Bristol, Bristol, United Kingdom; Division of Neurosurgery, Department of Surgery, National University Hospital, National University Health System, Singapore.
Division of Neurosurgery, Department of Surgery, National University Hospital, National University Health System, Singapore; Yong Loo Lin School of Medicine, National University of Singapore, Singapore.
World Neurosurg. 2022 May;161:291-302.e1. doi: 10.1016/j.wneu.2021.09.034.
Neurosurgeons today are inundated with rapidly amassing neurosurgical research publications. Systematic reviews and meta-analyses have consequently surged in popularity because, when executed properly, they constitute a high level of evidence and may save busy neurosurgeons many hours of combing and reviewing the literature for relevant articles. Meta-analysis refers to the quantitative (and discretionary) component of systematic reviews. It involves applying statistical techniques to combine effect sizes from multiple studies, which might offer more actionable insights than a systematic review without meta-analysis. Well-executed meta-analyses may prove instructive for clinical practice, but poorly conducted ones sow confusion and have the potential to cause harm. Unfortunately, recent audits have found the conduct and reporting of meta-analyses in neurosurgery (but also other surgical disciplines) to be relatively lackluster in methodologic rigor and compliance to established guidelines. Some of these deficiencies can be easily remedied through better awareness and adherence to prescribed standards-which will be reviewed in this article-but others stem from inherent problems with the source data (e.g., poor reporting of original research) as well as unique constraints faced by surgery as a field (e.g., lack of equipoise for randomized trials, or existence of learning curves for novel surgical procedures, which can lead to temporal heterogeneity), which may require unconventional tools (e.g., cumulative meta-analysis) to address. Therefore, it is also our goal to take stock of the unique issues encountered by surgeons who do meta-analysis and to highlight various techniques-some of which less well-known-to address such challenges.
如今，神经外科医生面临着大量神经外科研究出版物的冲击。系统评价和荟萃分析因此变得越来越流行，因为如果执行得当，它们代表了较高的证据水平，可以为忙碌的神经外科医生节省大量搜索和审查文献以寻找相关文章的时间。荟萃分析是系统评价的定量（和自由裁量）组成部分。它涉及应用统计技术来合并来自多个研究的效应大小，这可能比没有荟萃分析的系统评价提供更具操作性的见解。精心执行的荟萃分析可能对临床实践具有指导意义，但执行不当会造成混乱，并有可能造成伤害。不幸的是，最近的审计发现神经外科（以及其他外科学科）中的荟萃分析在方法学严谨性和对既定指南的遵守方面相对较差。其中一些缺陷可以通过更好地了解和遵守规定的标准来轻易纠正——本文将对此进行回顾，但其他缺陷源于源数据固有的问题（例如，原始研究报告不佳）以及手术领域面临的独特限制（例如，随机试验缺乏均衡性，或新手术程序的学习曲线存在，这可能导致时间异质性），这可能需要使用非常规工具（例如，累积荟萃分析）来解决。因此，我们的目标还包括评估进行荟萃分析的外科医生所遇到的独特问题，并强调各种技术——其中一些不太为人知——以解决这些挑战。
