Division of Biostatistics, University of Minnesota, Minneapolis, Minnesota, USA.
Statistical Research and Data Science Center, Pfizer Inc., New York, New York, USA.
Res Synth Methods. 2023 May;14(3):468-478. doi: 10.1002/jrsm.1626. Epub 2023 Feb 15.
A reference interval, or an interval in which a prespecified proportion of measurements from a healthy population are expected to fall, is used to determine whether a person's measurement is typical of a healthy individual. For a specific biomarker, multiple published studies may provide data collected from healthy participants. A reference interval estimated by combining the data across these studies is typically more generalizable than a reference interval based on a single study. Methods for estimating reference intervals from random effects meta-analysis and fixed-effects meta-analysis have been recently proposed and implemented using R software. We present an R Shiny tool, RIMeta, implementing these methods, which allows users not proficient in R to estimate a reference interval from a meta-analysis using aggregate data (mean, standard deviation, and sample size) from each study. RIMeta (https://cers.shinyapps.io/RIMeta/) provides users a convenient way to estimate a reference interval from a meta-analysis and to generate the reference interval plot to visualize the results. The use of this web-based R Shiny tool does not require the installation of R or any background knowledge of programming. We explain all functions of the R Shiny tool and illustrate how to use it with a real data example.
参考区间是指在一个特定的健康人群中,预期有特定比例的测量值落在这个区间内,用于判断一个人的测量值是否属于健康个体。对于特定的生物标志物,多个已发表的研究可能提供了来自健康参与者的数据。通过综合这些研究的数据来估计参考区间,通常比基于单个研究的参考区间更具有普遍性。最近提出并使用 R 软件实现了用于从随机效应荟萃分析和固定效应荟萃分析中估计参考区间的方法。我们提出了一个 R Shiny 工具 RIMeta,它实现了这些方法,允许不熟悉 R 的用户使用来自每个研究的汇总数据(均值、标准差和样本量),从荟萃分析中估计参考区间。RIMeta(https://cers.shinyapps.io/RIMeta/)为用户提供了一种从荟萃分析中估计参考区间的便捷方法,并生成参考区间图以可视化结果。使用这个基于网络的 R Shiny 工具不需要安装 R 或任何编程背景知识。我们解释了 R Shiny 工具的所有功能,并通过一个真实数据示例说明了如何使用它。