The Future of Proficiency Testing

In 1946 Sunderman learned the hard way that medical laboratories in Pennsylvania were not meeting their customers’ needs.  This was about 25 years before Crosby and his ground breaking definition of Quality (meeting customer requirements), but Sunderman understood that if clinicians felt compelled to send samples to multiple laboratories to get enough answers to collect and mean, then the laboratories were probably not generating credible values.

In order to sort out what was going on, he and his coinvestigator created a bunch of simulated samples and set them to the laboratories for testing.  Unfortunately what he discovered was that the clinicians were right, the laboratories were wrong and the laboratory information was crap strongly resembling a grand scale scattergram.  What followed was the development of a formalized Proficiency Testing scheme which eventually spread around the world.  

 That was a good thing because it introduced a new level of Quality assessment that has benefited laboratories greatly by making the processes of sample examination more rigorous and standardized.

 

But the world does not stand still and medical laboratory sample examination has become mechanized and computerized and quality controlled to the point that the machines are rarely wrong, and when they are they are smart enough to shut down reducing the risks of machine error.  So Proficiency Testing especially for machine generated laboratory data has become arguably redundant.   

Proficiency Testing for machine generated data has become a statistical exercise looking at things like bias and uncertainty, neither of which has much to do with laboratory proficiency and competence.  

That does not mean that there are no more laboratory errors; indeed there continue to lots of errors, most of which are the consequence of human foible such as some distraction, some mistake, some procedure or protocol error.  

And I strongly argue that human foible should be the focus of attention for Proficiency Testing going forward.

When we look at laboratory error, what we regularly see is that most reported errors are in the pre-examination phase, poor samples, wrong samples, insufficiency samples, contaminated samples.  In the examination phase we should be looking at least as closely at how the Quality Control was done as the testing outcome.  And in the post-examination phase we should be looking at things like what kind of report was created and how it was generated and to whom it was sent.  And there are other areas of focus including Quality Management procedures and Safety procedures and Transport procedures and Autoclaving practices.  

All these issues can easily be tested by Proficiency Testing methods.  In some situations they will be different from the traditional methods.  That is where the innovation and creativity becomes part of the play.  

So here is my warning:  Keep doing what we always do and Proficiency Testing will become progressively less relevant and more inappropriate and just plain irrelevant and wrong.  

Change or die.

More to come.