DysCo Lab, Department of Psychology, Université Paris Nanterre, F-92000 Nanterre, France
Laboratoire de Psychopathologie et Processus de Santé, Université de Paris, F-92100 Boulogne-Billancourt, France.
BMJ Ment Health. 2023 Feb;26(1). doi: 10.1136/bmjment-2022-300534.
Umbrella reviews are a new form of literature review that summarises the strength and/or quality of the evidence from all systematic reviews and meta-analyses conducted on a broad topic. This type of review thus provides an exhaustive examination of a vast body of information, providing the highest synthesis of knowledge. A critical strength of umbrella reviews is recalculating the meta-analytic estimates within a uniform framework to allow a consistent evidence stratification. To our best knowledge, there is no comprehensive package or software to conduct umbrella reviews.
The R package metaumbrella accomplishes this aim by building on three core functions that (1) automatically perform all required calculations in an umbrella review (including but not limited to pairwise meta-analyses), (2) stratify evidence according to various classification criteria and (3) generate a visual representation of the results. In addition, this package allows flexible inputs for each review or meta-analysis analysed (eg, means plus SD, or effect size estimate and CI) and customisation (eg, stratification criteria following Ioannidis, algorithmic GRADE or personalised classification).
The R package metaumbrella thus provides the first comprehensive range of facilities to perform umbrella reviews with stratification of the evidence.
To facilitate the use of this package, even for researchers unfamiliar with R, we also provide a JAMOVI module and an open-access, browser-based graphical interface that allow use of the core functions of the package with a few mouse clicks.
伞式综述是一种新的文献综述形式，它总结了广泛主题的所有系统综述和荟萃分析的证据的强度和/或质量。这种综述因此对大量信息进行了详尽的检查，提供了知识的最高综合。伞式综述的一个关键优势是在统一框架内重新计算荟萃分析估计值，以允许对证据进行一致分层。据我们所知，目前没有用于进行伞式综述的综合软件包或软件。
R 包 metaumbrella 通过构建三个核心功能来实现这一目标，这些功能包括 (1) 在伞式综述中自动执行所有必需的计算（包括但不限于两两荟萃分析），(2) 根据各种分类标准对证据进行分层，以及 (3) 生成结果的可视化表示。此外，该包允许对每个进行分析的综述或荟萃分析进行灵活输入（例如，均值加标准差，或效应量估计值和置信区间）和自定义（例如，按照 Ioannidis 的分层标准、算法 GRADE 或个性化分类）。
因此，R 包 metaumbrella 提供了第一个具有证据分层的伞式综述的综合功能。
为了便于使用该软件包，即使对于不熟悉 R 的研究人员，我们还提供了 JAMOVI 模块和一个开放访问的基于浏览器的图形界面，允许使用该软件包的核心功能，只需点击几下鼠标。
