Featured Post

Healthcare Customer Satisfaction: More Talk AND More Action

Healthcare Customer Satisfaction: More Talk AND More Action Customer satisfaction (Voice of the customer) is a recurrent th...

Monday, December 10, 2018

In-Country Proficiency Testing (EQA)



For those of you who live in the United States or Canada or Europe this will likely be a curiosity rather than the big deal that it actually is.  The North American history of Proficiency Testing started in the 1960-1980s when countries recognized the importance of externally monitoring the quality of medical laboratory test results.  We have in place very sophisticated systems.

The reality was, and still is, that in the vast majority of situations, laboratory test errors go undetected and a wrong result is reported to a physician or patient with frightening frequency. Even doing conventional quality control can’t control this.  The reason they go undetected is because when we go down, humans have enough internal mechanisms that we usually get better on our own regardless of what caused the problem.  Even if we have a diagnostic test that is ordered and performed with the best of intentions but is of wrong value, and is followed up with a medication that we don’t need, we usually have enough internal resources to tolerate all sorts of wrong impressions and inappropriate “medications” that we receive.  There are MANY people walking around with all sorts of missed or incorrect diagnoses. 
But every once in a while, a BAD thing happens, and a person either doesn’t get better, or gets worse and sometimes dies. By then it is too late, and all hell is likely to break loose.

So, our predecessors figured out that by requiring laboratories to test samples where they don’t know the actual result, but an external body does, we may catch more of these errors.  If the laboratory returns the correct accurate result, we inform them they are “proficient” or “competent”.  If they get it wrong, we don’t call them “incompetent” but harsh words hurt, but we say they have “opportunities for improvement”.  

The organizations that send out these quality check samples to laboratories are known either as Proficiency Testing bodies, or more frequently in most places around the world. External Quality Assessment bodies (or EQA).  

Now 30-50 years later, most wealthy countries have developed their own EQA bodies, that focus primarily on the quality concerns of their own countries.  Poorer countries have often struggled to get similar programs up on their own, and end up buying samples from the wealthy countries, which in the majority of situations has been a real failure approach.    

The samples that are sent all too often are inappropriate for the laboratories that receive them and importantly they don’t survive the travel, unless they have been freeze dried for the convenience of the sender (but not for the receiver!).  And more importantly, they were rarely designed with poor countries’ needs in mind.  They are designed for the complex sophisticated laboratories in the countries that created them.

We have struggled with this as a challenge for near 20 years with some success, but also a lot of failure (about 50:50).   We would invite countries to send staff to our EQA laboratory in Vancouver and would teach them a series of simple to learn, easy to implement methods that we developed so they could create their own stable samples at home.  

The methods are usually bullet proof even in climes with elevated temperature, but they depended up the country having some infrastructure to make the program happen.  And sometimes that infrastructure was not there.

So in a new collaborative program with another EQA provider, we are trying a new strategy.  Rather than have people come to us, we are going to them, working in their laboratories working with their staff, and ensuring at some essential parts of the necessary infrastructure, like the informatics system is in place.  So the samples can be made and be distributed in-country, and the information resulting can be simply entered into a computer hooked up to a sophisticated data base and the results can be sent to the key people.

Importantly, this way the country picks the challenges that it wants to send, with the microbiological test challenges it wants to use, rather than depending on samples from companies with a different approach.

So, this week we just did our first in-country training for in-country microbiology EQA in a county in West Africa.  The training program went even better than we anticipated, and plans are being put in place for the infrastructure and follow-up. 
Guaranteed success?  No… but so far things look really good and promising.  We will know a lot more in a few months from now.  But if it works the country will have taken control of the quality of its own laboratories, will have saved mountains of money by not buy samples from other countries, and will have the quality challenges that it needs, not what it receives.

Fingers crossed.


2 comments:

  1. For infectious disease testing, I would like to see the acceptable measurement of uncertainty applied at the medical decision point (grey zone)to ensure the accuracy of the assay instead of a 3sd peer group allowable limit. A "herd mentality" may not ensure the clinical value required to meet the customer's needs and requirements for patient care. The better laboratorians understand their assays performance and limitations, the better they can advise their clients.

    ReplyDelete
  2. Hi Anna
    Many thanks for your comment.
    The measurement of uncertainty, as much as it is incorporated into ISO15189, is a poor (indeed incorrect) measure of error and variance in the clinical setting for exactly the reasons that you are expressing.

    Uncertainly as a measurement tool was designed in Cern as a way to detect sound-from-noise in nuclear physics, where all parameters are under strict and absolute control.
    In the clinical laboratory virtually nothing is in control from before the sample was collected. We have no idea about inhibitors (medications, herbs) in the blood stream. We have poor control over transport, etc. so our uncertainty budget is MASSIVE. So applying MU to solely the examination phase component is really an exercise in folly.

    Fortunately in microbiology (including virology and serology as well as bacteriology, mycology and parasitology) there are almost no Quantitative studies, so MU does not apply. AT best our results are semi-quantitative, and again MU does not apply.

    Using MU as a method to monitor only the examination component only makes sense to accreditation bodies, and is pretty suspect.

    Ultimately the only real measures of performance are (1) correlation with clinical setting and (2) EQA (or maybe the other way around!!)

    M




    ReplyDelete