Meter Accuracy

Probably preaching to the choir with this post, but still:

I test twice with my meter every time I have a result >250. Sometimes the variations are huge, one example from last week: 428 mg/dl vs. 303 mg/dl. In this case, the deviation from the true value (which of course I don't know) has to be...

Please read over here, at my D-blog Pricked Finger in the Pie: Meter Accuracy

You wrote "Meters in Germany, as well as in the US, need to have a deviation of no more than 20% (only ~60% of the approved meters on the German market even fulfill this standard)".

This sentence is contradictory in itself. How can they not fulfill the standard they are regulated to fulfill? Which study came to the conclusion that about ~60% of the meters do not fulfill the regulation?

The "20%" accuracy requirement is only the "region A" of the Clarke Error Grid. In the US at least, meters are (a fraction of the time) allowed to be more than 20% off as long as they don't "lead to inappropriate treatment", e.g. region B on the Clarke Error Grid. So if your bg is really super high, the meter is allowed (at least some fraction of the time) to tell you that it is super high but a number different than 20% from the real value.

As a practical matter both "428" and "302" are (for me at least) too high. Also in my experience... those sorts of numbers come from having something sticky or gooey on my fingers. e.g. both the numbers are wrong and really my bg is likely to be more normal-ish!!!!

See e.g. Wikipedia Page on Clarke Error Grid

This picture shows typical bg meter performance, note that most all the readings outside region A (e.g. more than 20% off) are in region B.

Oh, thank you for pointing this out, the way I wrote it, it is contradictory indeed. Silly me. I'm going to add an Edit: to clarify. What I meant was this: Meters in the EU need to have an CE certificate, which is basically an statement by the manufacturer (or the importer), that the meter fulfills the standard European regulations. The certificate can be issued by a lot of so-called Notified Bodies in the EU. The fulfillment of the ISO norm 15197:2003 (deviations <20% or <15mg/dl for BS <75 mg/dl) is officially required for that, but nobody checks who checked for that (e.g. the manufacturer). A study by the university of Ulm lead by Dr. Freckmann and published in 2009 showed that actually only 60% of the meters in Germany fulfill the ISO norm.

Ok, I hope it's better now.

@Tim: A 300 is too high for my taste as well. But there was no gooey on my fingers and correcting for the 300 brought me smoothly back to 108, so think I interpreted correctly.
So, yeah, in the Clarke Error Grid the meter is of by more than 20% (regions B-E) in about 15% of the tests. That means every 6-7 test, or about once a day, depending on how often you test of course. I just fail to see how a deviation >20% is not affecting my treatment. Even if I'm at my aim of 100, a 25% deviation means my meter could show anything between 75 and 125, both numbers I would correct for in one way or the other. And the higher the blood glucose, the less deviation does it need to screw up my treatment.

I understand the desire for better meters.

But I grew up (and had diabetes) at a time when no meters were available anywhere.

I'm deathly afraid that if some congressperson starts insisting that bg meters have to be 1% accurate, that no meters will ever be able to hit that target (even lab machines do not get that sort of accuracy) and I'll find that the bg meters and test strips I depend on so much, will not be available. I remember the days of urine testing very very well. Wow, we do a million times better today.

In naval engineering and military circles, there's the phrase "better is the enemy of good enough".


Thanks for sharing were you're coming from. Sure, a bird in the hand is worth two in the bush. But I feel it wouldn't hurt to at least try improving meter accurcy rather than concentrating on less important stuff. If it doesn't work out, it doesn't work out, fine. But at least they could try.

Thanks for the clarification, Vera! The result of Dr. Freckmann's study is very important. It seems that the quality control of meters and test stripes is done poorly by the manufacturers. As you point out the risk to be held responsible for that or to loose market access is rather low. To me this is the root of the problem. I could accept 20% deviation if this standard is rigorously met. With my endos I had the chance to take more than 7 devices and to compare their results with lab equipment. The deviation I have seen was pretty high. For me the best results came from a device by the unknown manufacturer Terumo. It measures color changes like the good old test stripes from the 80ties and it has been developed in Japan.

I also started with the dreaded urine testing with Clinitest when I was diagnosed. When I think of it now, especially after reading what kids think of it today when they read this ancient history account of diabetes control, I have to agree with them that it's gross. I also have to say that if you ever want to drive yourself close to the border of losing your mind, try to find the best meter for accuracy. Happy Insanity.

Terumo? Sounds familiar... Oh, right, here it is: My lancing device is the FineTouch by Terumo. Best lancing device I ever had, as ugly as it may look, it's great, you should give it a try if you get the chance. </digression>