VerifyMyMU: nine top-down measurement-uncertainty models for medical laboratories in one offline HTML file

Abstract

Single-file, offline HTML application that computes the measurement uncertainty of a clinical measurement procedure from internal quality control and external quality assessment data under nine published top-down model formulations (ISO/TS 20914, ISO/TS 20914 with an added bias term, Nordtest TR 537 RMS and SEM forms, the mean-bias form, the EUROLAB TR 1/2007 reading, Cofrac SH GTA 14 and its range variant, ISO 21748). The four conventions the guidance leaves open — precision source, coverage rule, treatment of an uncorrected bias and source of u(Cref) — are separate settings, and an applicability screen classifies each model from the type of assigned value of the laboratory's EQA scheme by pre-specified criteria. Distributed with the verification scripts that reproduce the worked examples of Nordtest TR 537 ed. 4 and ISO/TS 20914 Table A.3, an independent Python re-implementation of the nine models, Monte Carlo coverage checks and user-interface tests. Research software; not a medical device. Accompanying paper: Kilinckaya MF, Nine top-down measurement uncertainty formulations on one clinical chemistry dataset, Scand J Clin Lab Invest (submitted).