Technical Report XXIX
Uncertainty Estimation of Diagnostic Accuracy Measures Under Parametric Distributions
RA Chatzimichail, T Chatzimichail, AT Hatjimihail
Abstract
Background: Diagnostic accuracy measures (DAMs) are widely used to evaluate medical diagnostic tests, but their uncertainty may affect threshold-based interpretation in clinical practice.
Methods: We developed a computational framework to estimate sampling uncertainty, analytical measurement uncertainty, and combined uncertainty for sixteen threshold-dependent DAMs under normal, lognormal, and gamma class-conditional measurement distributions. Analytical measurement uncertainty was represented using homoscedastic or heteroscedastic imprecision models, and uncertainty was propagated using a first-order Taylor-series approximation. The Wolfram Language program DiagAccU implements threshold-dependent estimation, uncertainty decomposition, confidence-interval calculation, and measure-specific threshold optimization.
Results: Fasting plasma glucose for diabetes diagnosis was used as an illustrative case study. Ratio-type measures showed marked instability near boundary thresholds, with widened confidence intervals, whereas bounded agreement, association, and concordance-based indices were comparatively more stable in interior threshold regions. Confirmatory objectives favoured higher thresholds emphasizing specificity, whereas exclusionary objectives favoured lower thresholds emphasizing sensitivity.
Conclusions: The framework provides a unified threshold-dependent approach for estimating uncertainty across multiple DAMs. It supports reproducible diagnostic-threshold evaluation, explicit uncertainty reporting, and clinically aligned interpretation for confirmatory diagnosis and diagnosis for exclusion.
First Published
2025-09-21
Revised
2026-08-13
DOI
Citation
Chatzimichail RA, Chatzimichail T, Hatjimihail AT. Uncertainty Estimation of Diagnostic Accuracy Measures Under Parametric Distributions. Technical Report XXIX. Hellenic Complex Systems Laboratory; 2025. https://doi.org/10.5281/zenodo.21921929.
Source Notebook (Revised on 2026-08-13)
Terms of Use
The material made freely available by Hellenic Complex Systems Laboratory is subject to its Terms of Use.