It appears you have not yet Signed Up with our community. To Sign Up for free, please click here....

Diabetes Message Board

Diabetes Board Index
Board Index > Diabetes | 0-9 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Your post is a little sloppy with use of the terms "accuracy" and "reliability"... which are NOT the same thing.

Now that you're aware that I know what I'm talking about ;) I'll move to the subject at hand: bG meters.

- - -
"20% or 20 points" is an often-discussed maximum target for future FDA approval requirements. In fact, no such 'standard' currently exists, and many current meters are FAR better than this when the bG value being tested is higher than 50 mg/dL. Many meter manuals, including mine, also present a 20% figure... And heck, if you're bG measures anywhere near 40 (or 50, or 30) who cares about the accuracy ??? You've got your diagnosis and your treatment plan: [B][I]Eat Sugar Tabs NOW !![/I][/B] 15 grams, wait 15 minutes, test again, and if necessary additional 15/15 sequences.

My strips and meter, OneTouch Ultra, are pretty good... but there are others which do approximately as well. The Owner's Booklet contains the usual chicken-s### statement "...value should agree with a laboratory measurement within +/-20% most of the time under normal conditions." But in fact, the Ultra is WAY better than 20%. The real performance figures are in the patient insert (present in every package of strips). For Accuracy versus the YSI model 2300 analyzer, (a really expensive and high quality machine), they got a correlation coefficient of .984 on 117 diabetic patients over a testing range of 36.4-434 mg/dL.

So overall, the accuracy was high: just a 2% chance of a test result being more than one Standard Deviation away from the YSI "true value". (So they could have claimed as little as "2%" instead of "20%" in the User Guide, because 1 S.D. of a normal distribution is considerably more than "most of the time", which I take to mean > 50% chance. But only for properly trained users, and they'd get zillions of inquiries from users with bad technique and users who didn't do the samples simultaneously, eating up profits. Thus the 20% claim, only someone who (a) seriously failed to follow instructions; or (b) had a busted meter or damaged strips, would have any realistic chance of doing worse than 20% with figures in the 70-350 range.)

The precision testing figures (how well the meter agrees with itself from strip to strip) had Coefficient of Variance values from 4.4% at low bG to 2.4% at high bG... using strips from different production runs (strips with different codes). When using strips from a single run (i.e., the same package), the corresponding figures were 3.2% and 1.6%.

And that's about what I get: if I test twice without delay, the two results are typically within 3-5 points of each other. I can't remember [B]ever[/B] seeing more than 8 points difference. BTW, I do this "two tests with no delay" very often, my Dexcom requires it for calibration of each new sensor, and I recommend it as a frequent meter check procedure (#6 of the right things to do, listed next).

- - - - -
There are a few things which I do right (If you're not getting results like mine, then you're not doing 'em):

(1) Never, never do "alternate site testing". The same patient insert includes has 9 warnings about arm testing, all of which say fingertip testing may show a significantly different result under many circumstances, and, in those circumstances, you should test (and trust the result) from your finger, NOT your arm.

(2) You should absolutely, positively, wash your hands before every test... and leave the water on them for a few seconds before you dry, so that any sugar left on your fingers from a morning coffee break donut (and etc.) is washed off before it contaminates your blood sample.

(3) Swing your arm around like a softball Pitcher, the centrifugal force gets a lot of fresh capillary blood down to your fingertips.

(4) Make a big drop, so that at least 1/3 of the blood is left behind after the strip is filled. And no squeezing allowed! Your sample will be polluted by interstitial fluid.

(5) Treat your strips right: store 'em in the house: a cool, dry, dark cupboard. The bathroom kinda sucks, the humidity is high, and the whole room can be warmed up a lot by showers and baths. If you're out in the Winter Cold, keep 'em in an inner pocket. Don't [B]EVER[/B] leave strips in a Glove Compartment and expect them to be OK!

A related issue: Depending on where you live, some strips just can't handle the weather. (That's the main reason I've been using OT-Ultra since it was invented, it's got the widest temperature range of any strip, 43F to 111F. I live in the Northern Nevada Desert.) The Precision XTRA, if they still sell it, sucks: only 64F to 86F! Even indoors, a little bit too much are conditioning and it's useless. And 86 isn't even body temperature! Here's the relevant URL at Children with Diabetes:


(6) Own and use more than one meter. Compare them to each other with every new jar of strips). If they disagree which each other, buy a new one, compare all three and throw away the bad one.

- - - -
Do these things with a decent meter strip combo (like mine), and your problems will be GONE. Like Coravh and me, you'll be looking at way under 10% CV, with correspondingly high Accuracy.

BTW, those "testing solutions" aren't STANDARD solutions, they're just a very poor tool to verify that the meter is more or less OK. My comparison procedure, #6, is the right way to go about it, it just costs an expensive strip instead of a cheap solution drop.
- - - -
Another BTW, for Mommyof4 of 9:50 PM on 5-30: All modern meters which I know of are made to present plasma-calibrated results, even though they use whole blood for the tests. (There might be exceptions, but I don't know of any. And it used to be different, 5 years ago many meters displayed whole blood figures in the way you say.) So Your MODERN meter and a lab machine such as the YSI should show nearly identical results for samples taken at the same time if you swing your arm around (per my #3) to assure fresh capillary blood for your test.

All times are GMT -7. The time now is 07:34 AM.

Site owned and operated by HealthBoards.comô
© 1998-2018 HealthBoards.comô All rights reserved.
Do not copy or redistribute in any form!