ABCMETAapp: R Shiny Application for Simulation-based Estimation of Mean and Standard Deviation for Meta-analysis via Approximate Bayesian Computation (ABC)

TL;DR

ABCMETAapp is an interactive R Shiny tool that estimates mean and standard deviation in meta-analysis from various summary statistics using approximate Bayesian computation, accommodating different outcome distributions.

Contribution

It introduces a flexible, user-friendly R Shiny application implementing ABC for estimating meta-analysis parameters from diverse summary data and distributions.

Findings

Supports multiple outcome distributions including Normal, Lognormal, Exponential, Weibull, and Beta.

Provides an interactive platform for practical implementation of ABC-based estimation.

Enhances estimation accuracy when original data are unavailable or skewed.

Abstract

Background and Objective: In meta-analysis based on continuous outcome, estimated means and corresponding standard deviations from the selected studies are key inputs to obtain a pooled estimate of the mean and its confidence interval. We often encounter the situation that these quantities are not directly reported in the literatures. Instead, other summary statistics are reported such as median, minimum, maximum, quartiles, and study sample size. Based on available summary statistics, we need to estimate estimates of mean and standard deviation for meta-analysis. Methods: We developed a R Shiny code based on approximate Bayesian computation (ABC), ABCMETA, to deal with this situation. Results: In this article, we present an interactive and user-friendly R Shiny application for implementing the proposed method (named ABCMETAapp). In ABCMETAapp, users can choose an underlying outcome distribution other than the normal distribution when the distribution of the outcome variable is skewed or heavy tailed. We show how to run ABCMETAapp with examples. Conclusions: ABCMETAapp provides a R Shiny implementation. This method is more flexible than the existing analytical methods since estimation can be based on five different distribution (Normal, Lognormal, Exponential, Weibull, and Beta) for the outcome variable.