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

Roopesh Reddy Sadashiva Reddy, Isildinha M. Reis+1 more|arXiv (Cornell University)|Apr 4, 2020

Scientific Computing and Data Management44 citations

TL;DR

This paper presents ABCMETAapp, an R Shiny app implementing ABC-based estimation of mean and SD for meta-analysis when only summary statistics are available, allowing different outcome distributions.

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.

Motivation & Objective

• Motivate the need to estimate means and standard deviations when primary studies only report partial summaries (e.g., medians, quartiles) for meta-analysis.
• Introduce ABCMETA, a simulation-based method to estimate mean and SD using approximate Bayesian computation.
• Provide an interactive R Shiny implementation (ABCMETAapp) enabling flexible distribution choices for skewed or heavy-tailed outcomes.

Proposed method

• Develop an R Shiny codebase implementing the ABCMETA method.
• Allow specification of multiple candidate distributions (Normal, Lognormal, Exponential, Weibull, Beta) for the outcome.
• Enable simulation-based estimation of mean and standard deviation from limited summary statistics.
• Provide an interactive workflow to run the method with example data.
• Offer user-friendly visualization and guidance within the Shiny app.

Experimental results

Research questions

• RQ1 Can ABC-based simulation accurately estimate mean and SD for meta-analysis using limited summary statistics?
• RQ2 How does allowing multiple underlying distributions (Normal, Lognormal, Exponential, Weibull, Beta) affect estimation flexibility and robustness?
• RQ3 Is the ABCMETAapp implementation user-friendly for practitioners performing meta-analyses with incomplete data?

Key findings

• ABCMETAapp provides a Shiny-implemented interface for the ABCMETA method.
• The method supports five distributions to model the outcome variable, accommodating skewness or heavy tails.
• The app demonstrates how to run the ABC-based estimation with examples.