ISO

In 2023-2024, Westgard QC conducted a survey of 1,300 laboratories around the world. When examining the various sizes of laboratories, mu use varies considerably. Who mu's the most? High volume or low volume labs? Who reports mu's the most? The answers willl surprise you...

Measurement Uncertainty Survey Results by Lab Volume:  One expected finding, and 5 unexpected ones. Insights into the

October 2024
Sten Westgard, MS

See all the Global MU Survey Results:

• The Global Practice of Measurement Uncertainty (minus the US)
• Europe sets the Standard
• The US stands alone
• Rising Asia, rising uncertainty
• Middle East, middling uncertainty
• Africa, where mu is new
• South and Latin America, a special case

• High Volume Labs
• Moderate Volume Labs
• Low Volume Labs
• Very Low Volume Labs

Methodology: From 2023-2024, Westgard QC conducted a global survey of laboratories about their use (or not) of measurement uncertainty. The labs were solicited by the Westgard website, the Westgard e-newsletter, advertising on Facebook and LinkedIn, as well as personal appeals across social media. More than 1,300 laboratories from more than 110 countries participated in the survey.

For this article, we're going to look at 4 different sizes of laboratories: High, Moderate, Low, and Very Low volume laboratories. In the graphs that follow, each lab type has the same color in every graph, i.e. high volume labs are always turquoise, moderate volume labs are always orange, etc.

Who calculate measurement uncertainty the most?

As might be expected, high volume laboratories calculate measurement uncertainty the most. But the drop-off as volume declines is not total. Even very low volume laboratories, almost half of them are calculating mu.

But what mu is being calculated?

As volume increases, the lab is more likely to embrace sophisticated versions of mu calculation. The lower the volume, the more likely the laboratory is going to use an easier alternative: the control SD as an estimate of mu.

Unfortunately for these labs, the godfather of mu, Mauro Panteghini, has insisted that using SD for mu is an incorrect approach. [See The simple reproducibility of a measurement result does not equal its overall measurement, Mauro Panteghini, Clin Chem Lab Med 2022; 60(10): e221–e222 DOI: 10.1515/cclm-2022-0618]

"[T]he MU at the level of clinical samples (u result ) must be the combination of all uncertainty contributions accumulated across the entire traceability chain. This refutes the common misconception...that the simple reproducibility of a measurement result equals its overall MU. A correct estimate of MU of laboratory results is indeed not possible without u cal, which in turn should include u ref."

Thus, we see that while a majority of labs are calculating mu, a majority of those laboratories are doing it wrong.

So how many laboratories don't know what measurement uncertainty they are calculating?

As lab volume increases, it appears that the ignorance of bias also increases. High volume labs are 10% more like to ignore the inclusion of bias in their measurement uncertainty than very low volume laboratories.

Do higher volume labs report mu less?

High volume labs might calculate mu the most, but they report it the least! Low volume and very low volumes labs were around twice as likely to report mu with their test results. That's an interesting development: those who are the biggest adopters are also the smallest reporters. Nearly all of these labs make mu available on request at the same rate.   

Do higher volume laboratories mondify their QC in response to mu?

Among the many actions one laboratory might take in response to mu, one routine option is to increase QC. But as the lab volume goes up, the changes to QC actually go down. Again, the higher the volume, the more a lab calculates mu, the less likely they are to make changes to QC in response to it.

Do higher volume laboratories have more requests for mu?

 The higher the volume, the less likely a lab reports that clinicians are asking for mu weekly or monthly. The labs that calculate mu the most, have clinicians asking for it the least.

Conclusion

As lab volume increases, labs are more likely to calculate measurement uncertainty more. But they are far less likely to use that mu, to report it, to have clinicians ask for it. Perhaps as labs grapple with more of the complications and implications of mu, they realize that it's not going to live up to the promises that the high priests of mu have been making.