Monday, August 13, 2012

Proficiency Testing in the News



It is always exciting to see stories in the media on topics that are of interest, even if the stories are not exactly positive.  It means that the shining light of public scrutiny approach to Quality is at work.

This story was first reported by a news affiliate in Columbus Ohio and was then widely communicated by Robert Michel in the Dark Daily [http://www.darkdaily.com/ebriefings#axzz23SJDJout].  It brings light to a prominent academic centre in the United States that has run into difficulties because they apparently “inadvertently” referred proficiency testing samples to another laboratory.  

Before going on, let me say that Dark Daily is one of the most valuable medical laboratory Quality oriented sites on the web.  It is highly informative and a must read.

The story does not give a lot of details, so I will describe the following discussion based on my own experiences rather than on the specifics of what did or did not happen in Ohio.

First of all let me start with the following bold statement.  Laboratorians as a collective have always shown an ambivalent attitude towards Quality Control, and Quality Assessment.  We know that QC and QA are critical components of ensuring confidence in laboratory performance, but that is pretty much tempered by an overwhelming libertarian nature that resents intrusions into our professionalism.  We may do QC and QA but we don’t like it. 

Many laboratorians have a flexible approach to demonstrating laboratory Quality Given a choice, many laboratories take the approach that the best Proficiency Testing program is the cheapest with the least number of samples and the most simplistic of challenges.  Others (I suspect and trust as not many) unfortunately go one step further taking it as far as they can through gaming and cutting corners.   You might call it being deceptive or maybe even dishonest.

In the “olden days” we used to regularly have laboratories that would send samples to reference laboratories for testing and then report the results as their own.  Even today, we have laboratories that don’t fill in forms, don’t adhere to deadlines, and quibble over ambiguities that they uniquely envision, and then gripe and complain.  It’s pretty much a “get out of our hair:” approach.  So with this as background, I can understand why officials at CMS would take a dim view towards finding laboratories that apparently are referring samples to other laboratories.  

In defense of the laboratory, I will point out that many academic centres have huge numbers of samples received every day, often reaching into the millions per year.  And life has become more complicated these days with organizations relying on transporting samples into specific regional centres rather than doing testing locally. So if a laboratory sends a sample or two  to a place it was not supposed to, the error rate would be pretty low (six sigma metric >5.5).  

But if they sent the sample to another laboratory, and then also reported the other laboratory’s results as their own, that is a big problem.

It is possible that there is another issue here; one where there is an absence of appropriate reporting choices.  The laboratory may be stuck with a form that they cannot fill in without creating another problem.

In our program we do have laboratories that receive our package of samples, which may contain challenges for certain tests that they would normally perform.   With CMPT our challenge menu is set annually and the package may include certain tests that are not normally performed by the laboratory.  For example they may do clinical bacteriology, but ship enteric samples (feces) to another laboratory.  They may do bacterial identification, but not susceptibility testing.  In those situations we cannot customize the packages, but we do not expect them to perform tests that they would normally not perform. 
 
We have a number of solutions.  The laboratory can complete and submit their report with the designation “SNNP” meaning “sample not normally processed”, which once confirmed results in an automatic “ungraded” sample.  

If they would normally do only a preliminary investigation and then send to another laboratory for completion, we make it clear that they provide us with the preliminary information that they generate and report that they would send it on. Clear instruction is given that they should not send the sample onward.
If they want to perform the challenge on an educational, self-interest, but not reported basis, they can also do that, with the result remaining as “ungraded”.  

Maybe these were options that the laboratory in Ohio did not have.

So at this point we don’t know if Dark Daily is reporting a problematic-shipping issue, or a caught-in-deception issue.  I hope for one and fear the other.

I will follow Dark Daily to see how the story unfolds.

2 comments:

  1. I'm a laboratorian and I resent your comments that we don't care about quality control or quality assurance. You're correct in that some activities we are required to perform are meaningless, but most of them are done for a purpose. The quality depends on the laboratory's leaders! The leaders must hire good people and ensure their training and competency.

    ReplyDelete
  2. I agree totally. Quality depends on the laboratory leaders and the laboratory culture they foster. Trained and competent workers working in laboratories that recognize the criticality of quality do not take short cuts or rob themselves of opportunities.

    ReplyDelete

Comments, thoughts...