R shiny app for estimation and verification of reference intervals using real-world laboratory data
10K+
A digital tool designed for the estimation and verification of reference intervals using real-world laboratory data. Built in R with the Shiny framework, the tool enables users to analyze up to 200,000 test results via simple copy-and-paste input - eliminating the need for file uploads or coding. The tool integrates multiple statistical methods, including refineR, TMC, TML, kosmic, and reflimR, and supports sex- and age-based stratification tools and dynamic drift detection algorithm. It provides automated data cleaning, visualization, and analysis. Sex-based differences are visualized using violin and boxplots, while statistical significance is assessed by non-parametric tests. Age stratification is guided by an iterative algorithm that merges age groups based on minimal median deviation within a user-defined number of bins, compared against dataset-derived uncertainty thresholds. The application is accessible via a public web interface and offers a Docker-based installation for secure, local deployment.
https://kc.uol.de/referenceranger/
Content type
Image
Digest
sha256:68c7492ec…
Size
584.1 MB
Last updated
7 days ago
docker pull gubruol/referenceranger