Lecture
Estimation of Reference Intervals - Better know your Limits
- at -
- ICM Saal 4a
- Type: Lecture
Lecture description
Reference intervals (RIs) and clinical decision limits form the foundation for the clinical interpretation of laboratory results. Current standards, including ISO 15189, require laboratories to verify and periodically review their own RIs. To support this process, a clear understanding of the nomenclature surrounding guide limits and the conceptual basis of reference intervals is essential. Direct methods represent the gold standard for determining RIs, relying on well-defined healthy reference populations. However, these methods are often infeasible in routine practice due to high resource demands, ethical limitations in vulnerable populations, and difficulties in assembling suitable non-diseased cohorts. Therefore, indirect methods - statistical algorithms applied to large sets of routine laboratory data - have become increasingly important. They not only enable laboratories to estimate their own population-specific RIs but also considerably simplify the verification of manufacturer-provided intervals.
This presentation elucidates the concept of guide limits and RIs, outlines the principles and relevance of direct and indirect methodologies, and highlights their role in meeting regulatory requirements. In this context a newly developed tool designed to facilitate RI estimation and verification using real-world laboratory data is introduced (ReferenceRangeR, https://kc.uol.de/referenceranger). The tool provides intuitive visualizations, statistical support for partitioning decisions, and implementation of multiple indirect RI estimation procedures (refineR, TMC, TML, kosmic, and reflimR). The source code of the application is available on GitHub and a Docker container enables a secure local deployment. The tool lowers the methodological and technical barriers and aims to help laboratories meeting regulatory standards and enhance the clinical reliability of laboratory diagnostics.
This presentation elucidates the concept of guide limits and RIs, outlines the principles and relevance of direct and indirect methodologies, and highlights their role in meeting regulatory requirements. In this context a newly developed tool designed to facilitate RI estimation and verification using real-world laboratory data is introduced (ReferenceRangeR, https://kc.uol.de/referenceranger). The tool provides intuitive visualizations, statistical support for partitioning decisions, and implementation of multiple indirect RI estimation procedures (refineR, TMC, TML, kosmic, and reflimR). The source code of the application is available on GitHub and a Docker container enables a secure local deployment. The tool lowers the methodological and technical barriers and aims to help laboratories meeting regulatory standards and enhance the clinical reliability of laboratory diagnostics.
