Tools, Technologies and Training for Healthcare Laboratories
  • Home

Consolidated measurement uncertainty specifications

In June 2022, the EFLM database began listing specifications for the extended allowable maximum measurement uncertainty. Now there is a formula to generate all the uncertainties you need..

This article is exclusively sponsored by LGC Technopath Clinical Diagnostics
Chasing the Mean - EQA, PT, and Peer Group

Consolidated Comparison of Maximum Measurement Uncertainty Performance Specifications

Sten Westgard, MS

Created June 13, 2022

Screen Shot 2022 06 13 at 3.43.17 PMUncertainty specifications are finally here. While measurement uncertainty (mu) has been mandated for decades, the benchmark for those measurements was equally uncertain. There was no consensus on how to calculate measurement uncertainty, and then once you did perform that calculation, how did you know if it was acceptable or unacceptable?

Previously, papers have proposed different types of uncertainty goals. There was permissible uncertainty (pu), and you can find a table of that listed here. There was (mupu)

The term du jour is now the Expanded Allowable Maximum Measurement Uncertainty (EAMMU, which for simplicity's sake, would be pronounced the same way as you pronounce, 'emu'). Actually, that's still not completely right, because these maximums are built on the minimum performance specification for imprecision, so it's more accurate to say EAMMMU. But if that is too much of a mouthful, you can resort to MAU, which is a different MAUthful altogether.

Below, we summarize the uncertainty goals. Where other goals than EFLM have been suggested, they are listed, but those are quite few. Please note this is not a full reproduction of the EFLM measurement uncertainty specifications, for that please visit the EFLM website directly.

In our other consolidated listings of performance specifications, we include a recommended choice.

In this case, we have no recommendations for mu goals, because we don't recommend measurement uncertainty. Full stop.

The third to last column is the EFLM allowable total error specificaton - again, now based on the minimum level of quality. This is calculated such that TEa will always be larger than EAMMU. However, that assumes that TE is being calculated such that only 2-3 SDs will fit into the TEa (in other words, aiming only for 2 or 3 Sigma level of quality). As regular visitors of the website - regular readers of Westgard publications - know, that is not the current recommendation. TEa is to be used as a specification for 6 Sigma. So any CV specification deriving from the Six Sigma approach using TEa is actually smaller and more demanding than that offered by EAMMU.

One will quickly note that the EAMMU is exactly twice the specification for CV. 2 SD, which sounds a lot like the practices of 50 years ago, when 2 SD control limits were the predominant, if not the only, QC procedure of the time.

Unfortunate to see so much effort poured into a recommendation that takes the laboratory backward, not forward, in quality.

 Measurement Uncertainty Performance Specifications
Test or Analyte Recommended Choice 2022 EFLM CV%
(minimum quality)
2022 EFLM
EAMMU
MAU

2022 EFLM TEa
(minimum quality)

2021 Panteghini et al pu
17-alpha-hydroxyprogesterone   <21.2% < 42.5%  ±53.0%  
25-hydroxy Vitamin D3   <5.2% <10.4%  ±18.7% <10.0%
Alpha-1 Antitrypsin   <3.1% <6.1%  ±9.3%  
Actual bicarbonate   <3.0% <6.0%  ±7.3%  
Adinoponectin   <14.1% <28.2%  ±43.7%   
Adrenaline   <101.3% <202.5%  ±221.6%  
Alanine aminotransferase (ALT)  
<7.6% <15.1% ±24.1% <4.65%
Albumin   <1.9% <3.8% ±5.2%  
Aldosterone   <27.5% <54.9% ±64.2%  
Alkaline phosphatase bone isoenzyme   <4.9%  <9.9% ±21.7%  
Alkaline phosphatase liver type   <4.0% <7.9% ±15.8%  
Alphafetoprotein   <3.4% <6.9% ±26.5%  
Amylase

 

<4.9% <9.9% ±19.8%  
Apolipoprotein A1   <4.1% <8.1%  ±11.3%  
Apolipoprotein B   <5.6% <11.1% ±17.2%   
Aspartate aminotransferase (AST)
 
<7.2% <14.4% ±20.5%  
Beta-2 microglobulin   <3.1% <6.1% ±9.7%  
Basophils   <9.3% <18.6% ±26.2%  
Bilirubin, total
 
<15.0%  <30% ±37.3% <10.5%
Calcitonin   <9.8% <19.5% ±41.2%  
C-Reactive Protein (CRP)   <25.6% <51.2%  ±76.1% <3.76%
C-Reactive Protein (hs-CRP)   <44.2% <88.3% ±109.4%  
Calcium, total     <1.4% <2.7% ±3.4% <0.91%
CA-125   <6.5% <13.0% ±20.3%  
CA 19-9   <3.2% <6.4% ±26.9%  
CA 72-4   <37.7% <75.4% ±105.4%  
Carbon dioxide total   <3.0% <6.0% ±7.3%  
CEA   <5.1% <10.2% ±30.8%  
Chloride   <0.8% <1.7% ±2.0% <0.49%
Cholesterol, total   <4.0% <7.9% ±13.0%  
Cholesterol, HDL    <4.3% <8.7% ±16.6%  
Cholesterol, LDL   <6.2% <12.5%  ±20.5%  
Chromograinin A   <11.1% <22.2% ±31.1%  
Complement C3   <3.4% <6.9%  ±11.6%  
Complement C4   <5.2% <10.4%  ± 18.1%  
Copper   <5.6% <11.3% ±15.0%  
Cortisol   <12.2% <24.5%  ±39.4%  
Creatine kinase    <11.3% <22.5% ± 33.8%  
Creatinine    <3.4% <6.8% ±11.1% <2.2%
Cystatin C   <3.0% <6.0% ±9.7%  
Eosinophils   <11.3% <22.5% ±43.7%  
Erythrocytes   <2.0% <3.9% ±5.8%  
Estradiol   <11.3 <22.5% ±26.0%  
Follicle stimulating hormone FSH   <9.3% <18.6% ±31.8%  
Free kappa light chain   <3.6% <7.2% ±12.0%  
Free lambda light chain   <3.6% <7.2% ±12.7%  
Fructosamine   <1.7% <3.4% ±5.4%  
GGT   <6.8% <13.6% ±28.3%  
Glucose   <3.8% <7.5% ±9.8% <2.0%
Glycated Albumin   <1.0% <2.1% ±3.9%  
Haptoglobin   <6.4% <12.9% ±25.6%  
HbA1c (NGSP)   <0.9% <1.8%  ±3.3% <3.0%
HbA1c (IFCC)   <1.2% <2.4% ±4.7%   
Hemoglobin   <2.0% <4.1%  ±5.8% <2.8%
Hematocrit   <2.1% <4.2% ±5.8%  
IgA   <4.3% <8.6%  ±14.7%  
IgG   <2.6% <5.3%  ±10.9%  
IgM   <4.4% <8.9%  ±25.6%  
Insulin   <19.0% <38.1% ±47.2%  
Iron, total   <15.5% <31.0% ±40.0%  
Lactate dehydrogenase (LDH)   <3.9% <7.8% ±11.5%  
Leukocytes   <8.1% <16.2% ±20.7%  
Lipase   <6.9% <13.8% ±21.3%  
Luteinising hormone (LH)   <17.1% <34.2% ±42.6%  
Lymphocytes   <8.1% <16.2% ±22.8%  
Magnesium   <2.2% <4.3% ±6.0%  
MCHC   <0.8% <1.5% ±1.9%  
MCH   <0.6% <1.2% ±2.7%  
MCV   <0.6% <1.2% ±2.4%  
MPV   <1.7% <3.4% ±5.6%  
Monocytes   <10.0% <20.0% ±26.2%  
Neutrophils   <10.5% <21.0% ±27.6%  
Parathyroid hormone   <11.8% <23.5% ±30.0%  
Phosphorous   <5.8% <11.7% ±14.6%   
Plateletcrit   <4.8% <9.6% ±13.5%  
Platelet distribution wide (PDW)   <2.8% <5.7% ±9.5%  
Potassium    <3.1%  <6.1% ±7.3% <1.96%
Prolactin   <22.1% <44.3% ±56.1%   
PSA   <5.1% <10.2% ±24.4%  
PSA, free   <5.3% <10.6% ±26.3%  
Protein, total   <2.0% <3.9% ±5.2%   
Reticulocytes   <7.3% <14.5% ±22.8%  
SHBG   <7.3% <14.5%  ±25.8%  
Sodium   <0.4% <0.8% ±1.0% <0.27%
Testosterone   <9.4% <18.8% ±24.8%  
Testosterone, free   <16.5% <33.0%  ±40.9%  
Thrombocytes   <5.7% <11.4% ±17.0%  
Thyroglobulin   <7.9% <15.9%  ±42.3%  
Thyroid stimulating hormone (TSH)   <13.3% <26.5% ±36.9%  
Thyroxine free  (FT4)   <3.7% <7.4% ±9.5%  
Thyroxine total (TT4)   <4.8% <9.6% ±13.0%  
Total tri-iodothyronine (T3)   <7.1% <14.1% ±17.4%  
Triiodothyronine (FT3)   <3.8% <7.5% ±9.8%  
Transferrin   <2.9% <5.8%  ±10.2%  
Triglycerides     <15.0% <30.0% ±40.6%  
Troponin I   <9.0% <18.0% ±29.0%  
Troponin T   <8.6% <17.1% ±26.5%  
Urea Nitrogen   <10.4% <10.9% ±26.6% <7.05%
Uric acid   <6.2% <12.5% ±19.2%  
Vitamin A   <4.6% <9.1% ±15.7%   
Vitamin E   <5.3% <10.6%  ±15.8%  
Zinc   <6.4% <12.8%  ±15.1%  

Here are the references:

EFLM database, accessed 6/13/2022. https://biologicalvariation.eu/meta_calculations

Performance specifications for measurement uncertainty of common biochemical measurands according to Milan models, Federica Braga and Mauro Panteghini, CCLM 2021; 59(8):1362-1368.


 

Joomla SEF URLs by Artio