Personal

Starting out the year on positive note? Or should you focus on the negative instead...

New Year, New Goals, and New Errors

Or, why we need to stay focused on errors even as we near their extinction

January 2019
Sten Westgard, MS

My name is a series of errors. My last name, the well known “Westgard” should rightly be Westgård with an å, known as an overring above the A. When you translate “Westgård” from Norwegian, you get “West Farm” which is where the family farm was settled out in Plaza, North Dakota. But under the crude millstone that is the English language, many nuances are ground down, and thus we have “Westgard.”

My first name, Sten, is also an error. Originally, I was going to be called “Sven” which is a far more common Scandinavian name. But my older sister, who was but a few years old when I was born, was presented with this name, she mispronounced it “Seven.” My parents didn’t think that having a sister pronounce my name as a number was a good sign, so they changed it to something that was close, but easier for my sister to pronounce. (“Sten” in most Scandinavian languages, translates literally to “Stone”, which may also explain my stubbornness.)

My middle name is Anders, another Scandinavian name that’s close to the equivalent of “Andrew”, but not quite. I guess the thinking was, if my strange first name wasn’t going to work out for me, then a strange middle name would be my back-up plan. The chance that someone in the US can pronounce “Anders” correctly when they first see it is about as low as the chance that they can pronounce “Sten” correctly. In the US, I am constantly called Stan, Seth, Steve, any combination of the four letters that is anything but what my name truly sounds like. Over the years, I have become accustomed used to answering to a lot of names that sound close to my real one. Only in Scandinavia do I have the pleasure of hearing my name exquisitely pronounced. (In defense of my parents, they were very proud of giving their children authentic Scandinavian names, and I don’t begrudge them for it)

But you see, with a name that is really just a series of errors, I hope that helps explain why my career’s focus is on defects. My concentration is on the failures that bedevil laboratory testing, rather than the successes.

There are times when it seems like the focus on errors is too pessimistic, too much of a “downer”. Unequivocally, we have come so far in healthcare and diagnostics. We produce miracles today that were unimaginable only a few years ago. Methods continue to improve. Automation, informatics, and productivity are surging.

So why do we insist on complaining in a time of plenty? Because that time of plenty is only available to a tiny few who enjoy access to the best laboratories with the best equipment and the best skilled staff. Because even as we enjoy these technological advances, the economic and political forces surge in the opposite direction, threatening to erode the foundation of medical diagnostics so severely, the edifice may no longer stand. Because as we introduce each new miracle, we struggle to adapt our quality systems, so much that our current hold on the quality of all of our methods may be very tenuous indeed.

The threats to our progress are manifold. New technologies in the “omics” sphere raise the potential benefits -- and the possible errors -- by orders of magnitude. When we’re trying to monitor millions and billions of results, our capability to monitor and assure quality may be severely limited at best. Right where the results are most definitive and life changing, we may not have proper quality control at all. The accelerating proliferation of point-of-care devices threatens to drown the public in rapid, easy, and inaccurate results, degrading the overall perception of laboratory tests. And the entrance of new diagnostic manufacturers producing low quality (but cheap) instrumentation poses that same threat within the core laboratory.

The threats are also fundamental to our quality system. Our Quality theories and models are in conflict and the official imprimateur of ISO 15189 has firmly placed its weight on the side of measurement uncertainty. Measurement uncertainty claims to be the most scientifically superior, methodologically pure, and mathematically correct technique to determine how bad your tests results are. The “Total Error” approach, adapted more recently to the Six Sigma scale, is a practical way to determine how good your test results are. With Sigma-metrics, you can determine how good your method is, how many “Westgard Rules” are necessary, how many controls are needed, and even how often you need to run those controls. Nearly every practical question about quality control can be answered with the Sigma-metric approach, while none of those questions can be answered with measurement uncertainty. Where measurement uncertainty is most useful, ironically, it is least applied: adding it to the test report for the clinician to evaluate and place test results in context. Few laboratories use it that way – most simply calculate it out of obedience to their compliance needs, but avoid using mu in any reports.

Finally, we are seeing a move toward greater control and centralization of analytical performance specifications. While the “Ricos group”, an informal set of colleagues, managed the database, they produced an update to the biological variation database every 2 years for more than a decade (1999 through 2014). Now that the EFLM has acquired the database, it has been more than 4 years since the data was updated, and when it will be updated, it sounds as if the information will be surrounded by banners, ads and corporate sponsorships. EFLM has effectively monetized the database, something that wasn’t done for the first 15 years of its existence.

That’s one of the reasons why we’re continuing to present the 2014 database entirely free of commercial sponsorship here on Westgard Web. And we are also posting comparisons of the new EFLM performance specifications with the other global goals that are available. The best use of these goals is for them to be disseminated widely, so everyone has a chance to evaluate and implement them. If the goals are hidden behind a complex login, or a jungle of commercials, that will make them that much harder to access and use.

We hope these new goals will help you on your lab’s journey toward success in 2019 and beyond.