MEASUREMENT UNCERTAINTY
Setting measurement uncertainty limits: a brief introduction to the issues
In this third article in his new series, Stephen MacDonald summarises some interesting highlights of the current understanding of analytical performance specifications and in particular how they are applied to measurement uncertainty.
In this next of the series of articles focused on development in the application of measurement uncertainty, we discuss probably the most common question that is asked about MU. That question being – what are the limits for MU? This is a very challenging question to answer and has been a source of debate in the peer- reviewed literature for decades. Billed as the most important meeting you didn’t
know was happening, I couldn’t wait to go to the 5th symposium of the Cutting Edge of Laboratory Medicine in Europe (CELME) meeting in Prague in October 2023.
It was clear from the outset that the intention of this meeting was not to revise the Milan Criteria (which have been referred to in previous articles, across multiple series!) for analytical
performance specifications (APS), but rather to act as more of a review of how those models had been applied in the preceding decade. Rather than review the entire topic of APS, I have summarised some interesting highlights of the current understanding of APS and in particular how it is applied to the measurement uncertainty method.
Development of models required for APS Standardisation of methods requires confirmation of commutability of reference materials to patient samples. In doing so an established calibration hierarchy can be determined and results from such methods can be used interchangeably. Central to this is agreeing limits of acceptable performance so that uncertainty propagated through this traceability chain does not exceed the maximum amount allowed before clinical results are impacted.
1999 – IFCC/
Tonk’s rule in 1963 and onwards
CAP Aspen conference
IUPAC conference (Stockholm)
2014 – EFLM Milan conference
Analytical Performance Specifications
are targets for performance, applied to an individual measurand that is determined by the clinical utility of the result.1
Accurate determination of
APS is essential not only for routine monitoring of measurement procedures in clinical laboratories but also to provide performance limits for EQA schemes and guidance for IVD manufacturers about expected performance of assays. The Milan consensus agreement2
1970s Cotlove, Harris and Williams
1992-1995 – EGE Lab and EQA organisation working groups
2014 – EQALM symposium
2023 – CELME meeting, Prague
from the EFLM/IFCC refined the models initially presented at the first IFCC/IUPAC conference in Sweden in 1999,3 by Fraser in 2015.4
However, although
Fig 1. The timelines of Analytical Performance Specifications meetings, starting from the original rules of consensus from Tonks in the 1960s to the most recent iteration in Prague, 2023.
WWW.PATHOLOGYINPRACTICE.COM APRIL 2024
considered the first formalised meeting on APS in 1999, the discussion of the requirements of APS began many years before with Tonk’s rule in 1963 (Fig 1).5 In the context of MU, setting
29 reviewed
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57
