Accreditation breaks on the hard rock of EQA

External Quality Assurance is intended to reveal where the best intentions of good practice fall down.  It shows how individual laboratory’s numerical results compare when examining a batch of identical samples sent to multiple labs.

Fortunately, it also exposes accreditation’s inability to assure consistent results between labs, although no explicit comparison between accredited and non-accredited labs has been done.  Nor has anyone investigated whether individual labs’ performance can be improved over time by accreditation.  That is a very curious omission considering a) the legal and commercial coercion to be accredited, and b) the understanding the medical staff have of assessing the strength of evidence in clinical trials.

Victims of ISO 9001 who are honest know that it doesn’t work.  They’ve known it for a long time.  This is why the cartel member that has been allocated your country tries to get labs to advertise itself on victim company’s literature by telling them they are promoting their own validity.  The ISO standard wants victims to demand accreditation of any other company they deal with because, otherwise, who would bother with it?  This is called coercion.  In many other contexts it is illegal.

That's why they developed the laboratory ISO standards, Dilbert.

EQA results from different labs examining the same samples show that attempts to improve ISO 9001 to assure correct measurements are flawed .  Getting your hands on EQA data can be hard if you’re not a scheme member or not motivated to ask for it in the public interest under the Freedom of Information Act. Not so long ago there were investigative journalists who did more than cut and paste press releases.  So try these collations of results from many microbiology laboratories to see the magnitude of the problem.

Some results can be pretty good for a distribution of theoretically identical samples:

Variation between labs of around 1 log unit

Did you think accreditation would mean there would be almost no variation between the values determined by different labs?  Has this changed is the confidence UKAS sold you?

Others can show much wider variation:

Variation between labs of over 3 log units!

Several labs are wildly out.  Has accreditation actually been improving lab’s performance over the years?  Why has no-one bothered to investigate?  Without testing, how would you know that accreditation doesn’t make things worse?

Imagine what might be revealed if someone undertook an meta-analysis of EQA results.

Confidence?  There is no evidence that accreditation gives assurance of making results any better after all.

And yet the EU uses supposed consistency between laboratories as the reason to enforce ISO 17025 on the Official Control Laboratories it forced on member states.  French law is requiring clinical labs to work to ISO 15189.  Such things don’t happen merely by consensus between gullible peers but though influential, high-level personal contacts between endemically corrupt eurocrats.

I wonder where this will end up?

But EQA itself is not the final word either.  Like accreditation, its efficacy has not been proven.  Some lack of certainty remains.  Sten Westgard has observed that in clinical chemistry,

“-most EQA programs and even peer-group programs – are not based in accuracy. Instead, those surveys are only based on “consensus.”….

“Another way of looking at this issue: this is the retail end of Traceability. While there are committees of metrologists talking about the importance of standardization, harmonization and traceability, their discussions can seem a little abstract. Accuracy-Based Surveys offer labs a chance to quantify the biases affecting their own methods – the real-world effects of a lack of standardization. And ultimately, it will be pressure from laboratories – spurred by these uncovered biases found in Accuracy-Based Surveys – that drive manufacturers to make real improvements in their methods.”

So just because you don’t understand what the accreditors are saying doesn’t mean they’re correct.