Center for Clinical Trials, Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland 21205, USA.
Ophthalmology. 2013 Jun;120(6):1113-9. doi: 10.1016/j.ophtha.2012.11.038. Epub 2013 Mar 22.
Systematic reviews and meta-analyses serve as a basis for decision-making and clinical practice guidelines and should be carried out using appropriate methodology to avoid incorrect inferences.
We describe the characteristics, statistical methods used for meta-analyses, and citation patterns of all 21 glaucoma systematic reviews we identified pertaining to the effectiveness of prostaglandin analog eye drops in treating primary open-angle glaucoma, published between December 2000 and February 2012.
We abstracted data, assessed whether appropriate statistical methods were applied in meta-analyses, and examined citation patterns of included reviews.
We identified two forms of problematic statistical analyses in 9 of the 21 systematic reviews examined. Except in 1 case, none of the 9 reviews that used incorrect statistical methods cited a previously published review that used appropriate methods. Reviews that used incorrect methods were cited 2.6 times more often than reviews that used appropriate statistical methods. We speculate that by emulating the statistical methodology of previous systematic reviews, systematic review authors may have perpetuated incorrect approaches to meta-analysis.
The use of incorrect statistical methods, perhaps through emulating methods described in previous research, calls conclusions of systematic reviews into question and may lead to inappropriate patient care. We urge systematic review authors and journal editors to seek the advice of experienced statisticians before undertaking or accepting for publication a systematic review and meta-analysis.
FINANCIAL DISCLOSURE(S): The author(s) have no proprietary or commercial interest in any materials discussed in this article.
系统评价和荟萃分析可作为决策和临床实践指南的基础，应采用适当的方法进行，以避免错误的推断。
我们描述了 21 篇有关前列腺素类似物滴眼剂治疗原发性开角型青光眼疗效的青光眼系统评价的特征、荟萃分析中使用的统计方法和引用模式，这些评价发表于 2000 年 12 月至 2012 年 2 月之间。
我们提取数据，评估荟萃分析中是否应用了适当的统计方法，并检查了纳入的综述的引用模式。
我们在检查的 21 篇系统评价中有 9 篇发现了两种有问题的统计分析方法。除了 1 种情况外，在 9 篇使用错误统计方法的综述中，没有一篇引用了使用适当方法的先前发表的综述。使用错误方法的综述被引用的次数比使用适当统计方法的综述多 2.6 倍。我们推测，通过模仿先前系统评价的统计方法，系统评价作者可能延续了错误的荟萃分析方法。
使用错误的统计方法，可能是通过模仿先前研究中描述的方法，使得系统评价的结论受到质疑，并可能导致不适当的患者护理。我们敦促系统评价作者和期刊编辑在进行或接受系统评价和荟萃分析之前，寻求有经验的统计学家的建议。
作者在本文讨论的任何材料中均无专有或商业利益。
