Meta-DiSc 2.0: a web application for meta-analysis of diagnostic test accuracy data

Area of Science:

• Biostatistics
• Medical Informatics
• Epidemiology

Background:

• Systematic approaches are crucial for estimating diagnostic test accuracy (DTA).
• Statistical meta-analysis of DTA studies presents challenges for non-experts.
• The original Meta-DiSc software (2006) facilitated DTA meta-analyses but required updates.

Purpose of the Study:

• To develop an updated, user-friendly version of Meta-DiSc software.
• To incorporate advanced statistical methods, including hierarchical models.
• To enhance the web-based interface for improved user experience.

Main Methods:

• Developed Meta-DiSc 2.0 as a web application using R Shiny.
• Implemented state-of-the-art bivariate and univariate random effects models.
• Included comprehensive analyses of heterogeneity, subgroup, and meta-regression.

Main Results:

• Meta-DiSc 2.0 provides a bivariate random effects model for DTA meta-analysis.
• The application features an intuitive interface with graphical and analytical tools.
• Algorithms were validated against established statistical packages (STATA/SAS, MetaDTA).

Conclusions:

• Meta-DiSc 2.0 is an updated, accessible, and statistically robust software for DTA meta-analysis.
• The new version improves upon previous limitations with advanced statistical models.
• The freely available web application aims to support researchers in DTA reviews.