ISTH Congress Abstracts

Official abstracts site for the ISTH Congress

MENU 
  • Home
  • Congress Archive
    • ISTH 2022 Congress
    • ISTH 2021 Congress
    • ISTH 2020 Congress
  • Resources
  • Search

Estimation and Analysis of Measurement Uncertainty Acceptability: The Interest of Bayesian Inference

F. Sobas1, K. Bourazas2, M.O. Geay3, M. Beghin4, E. Jousselme5, C. Nougier5, P. Tsiamyrtzis6

1Hospices Civils de Lyon, Hemostasis Laboratory, Bron, France, 2Athens University of Economics and Business, Athens, Greece, 3Hospices Civils de Lyon, Pierre Bénite, France, 4Hospices Civils de Lyon, Lyon, France, 5Hospices Civils de Lyon, Bron, France, 6Politechnico di Milano, Department of Mechanical Engineering, Milan, Italy

Abstract Number: PB0521

Meeting: ISTH 2020 Congress

Theme: Diagnostics and OMICs » Laboratory Diagnostics

Background: According to the ISO 15189 standard, section 5.5.1.4, “The laboratory shall determine measurement uncertainty (MU) for each measurement procedure in the examination phase used to report measured quantity values on patients’ samples. The laboratory shall define the performance requirements for the measurement uncertainty of each measurement procedure and regularly review estimates of measurement uncertainty.” Section 5.5.1.2 of the standard also specifies that “The laboratory shall obtain information from the manufacturer/method developer for confirming the performance characteristics of the procedure without modification.” From this point of view, Bayesian predictive inference is an especially well-adapted approach for IQC management, as several reports have made clear.

Aims: Taking factor-II assay as an example, the present study aimed to explain the usefulness of Bayesian inference applied to IQC results analysis, as a means both of calculation and of assessing the acceptability of estimated MU.

Methods: The Bayesian approach provides a predictive distribution of the future observable. This sequentially updated distribution can be used to determine a region where the future observable will (most likely) be, as long as no special causes are present. Using the predictive distribution, a 100(1-a) % Highest Predictive Distribution (HPrD) region, defined by the lower and upper PCC limits (figs 1 and 2), estimates MU with 95% probability if the chosen a value is 5 % (Tsiamyrtzis et al., Blood Coagulation and Fibrinolysis 2015; 26:590-6).

Results: The manufacturer´s maximum acceptable coefficient of variation defines the maximum acceptable MU in terms of lower and upper range limits (fig. 1). In truly efficient laboratories, MU is significantly lower (fig. 2).

Conclusions: Bayesian inference is an appropriate means of estimating measurement uncertainty for methods not subject to modification.


[Figure 1: Bayesian estimation of uncertainty of measurement with the range Lower PCC limit – Upper PCC limit with respect to a minimal performance]


[Figure 2: Bayesian estimation of uncertainty of measurement with the range Lower PCC limit – Upper PCC limit with respect to a optimal performance]

To cite this abstract in AMA style:

Sobas F, Bourazas K, Geay MO, Beghin M, Jousselme E, Nougier C, Tsiamyrtzis P. Estimation and Analysis of Measurement Uncertainty Acceptability: The Interest of Bayesian Inference [abstract]. Res Pract Thromb Haemost. 2020; 4 (Suppl 1). https://abstracts.isth.org/abstract/estimation-and-analysis-of-measurement-uncertainty-acceptability-the-interest-of-bayesian-inference/. Accessed October 1, 2023.

« Back to ISTH 2020 Congress

ISTH Congress Abstracts - https://abstracts.isth.org/abstract/estimation-and-analysis-of-measurement-uncertainty-acceptability-the-interest-of-bayesian-inference/

Simple Search

Supported By:

Takeda logo

ISTH 2022 Congress site

Visit the official web site for the ISTH 2022 Virtual Congress »

  • Help & Support
  • About Us
  • Cookies & Privacy
  • Wiley Job Network
  • Terms & Conditions
  • Advertisers & Agents
Copyright © 2023 John Wiley & Sons, Inc. All Rights Reserved.
Wiley