This 15% rule is just plain ridiculous

Glucoweb · #1 (**permalink**) 03-26-2008, 07:38 AM

I know the guidelines for BG Meters is that are required to by accurate +/- 15%. OK, I just took 3 readings in row, using different fingers. My readings were in this order:
READING A: 207
READING B: 179
READING C: 200

- OK, -15% Of the Reading A gives me 175.95. The other two readings fall within the accuracy guidelines.

- BUT, +15% of Reading B gives me 205.85 not within the accuracy guidelines.

- All three are within 15% of Reading C (-15% = 170 and +15%= 230)

I really wish the FDA would titghten down this 15% to at least 7.5% accuracy guideline.

Scratch · #2 (**permalink**) 03-26-2008, 07:41 AM

I thought it was +/- 20% which probably won't make you any happier regarding the somewhat diverse readings you had above, although the 207 and 200 would likely indicate your true glucose level is closer to that than 179.

I don't know. I think it's something people tend to fret too much over and in some ways I don't want to give the manufacturers any more potential reasons to jack up the costs on the testing strips.

morrisma · #3 (**permalink**) 03-26-2008, 07:48 AM

Although I seriously doubt one meter is significantly better than any others, there are ample opportunities to try them out for free. I have seen at least one free meter offer a month and they always come with enough strips to do an accuracy test such as the one you describe. The downside of course is that you need to keep doing the test on every lot until you are sure they can (or cannot) meet some higher standard.

There may also be online results of more scientific testing on meter's websites and maybe independent sites as well.
Mike

johgn · #4 (**permalink**) 03-26-2008, 08:35 AM

I was under the impression you would need to use the same finger each time for this test.

princesslinda · #5 (**permalink**) 03-26-2008, 08:51 AM

Quote:

Originally Posted by johgn

I was under the impression you would need to use the same finger each time for this test.

Even using the same finger, same stick, I have yet to get 2 readings in a row that were exactly the same.

xMenace · #6 (**permalink**) 03-26-2008, 09:23 AM

Quote:

Originally Posted by princesslinda

Even using the same finger, same stick, I have yet to get 2 readings in a row that were exactly the same.

I have!

But only once, ever.

I am planning on running a test sometime of various meters using calibration solutions. My daughter is studying chemistry at university, and I'm hoping she'll prepare some glucose solutions at various levels for me to use. I'd like to know how close the very high and very low readings are. I may just talk to her prof or another one directly. It should be easy to do. I also notice that sanvita produces test solutions in three strengths. I may try to get some of those.

Scrabblechick · #7 (**permalink**) 03-26-2008, 09:51 AM

Good idea, X! Let us know what you come up with.

I've said before that I have two meters, a True Track and a One Touch Ultra2. The One Touch consistently runs about 5-10 points higher than the TT. When it's not 5 points lower. LOL. Yesterday, I did three consecutive tests with my TT and the results were within 5 points of each other. So, I'm pretty comfortable with its accuracy and consistency.

I just have to use one meter and stick with it, or my fingers would leak blood! LOL. My One Touch is my backup meter.

As long as my next bloodwork agrees with the results I've been getting on my TT, I'll be a happy camper!

xMenace · #8 (**permalink**) 03-26-2008, 09:56 AM

I always get my math wrong, but here's something to consider

(179 + 207) / 2 = 193 (estimated median)

193 * 1.15 = 222 (+15%)
193 * .85 = 164 (-15%)

179 / 193 = 93%
207 / 193 = 107%

Your numbers seem ok to me.

Glucoweb · #9 (**permalink**) 03-26-2008, 10:15 AM

Quote:

Originally Posted by xMenace

I always get my math wrong, but here's something to consider
(179 + 207) / 2 = 193 (estimated median)
193 * 1.15 = 222 (+15%)
193 * .85 = 164 (-15%)
179 / 193 = 93%
207 / 193 = 107%
Your numbers seem ok to me.

I do not believe the quidelines discuss taking the median of multiple readings then make sure all the readings fall within 15% of the median. I did find some interesting information here:
Diabetes Information - Glucose Meters and Diabetes Management
I like the 1993 recommendation, too bad it was never followed and that was 15 years ago.

Lizzie G · #10 (**permalink**) 03-26-2008, 10:25 AM

one time i got a reading of 2.7 (around 47 in U.S units). i thought WEIRD, lowest ever reading and i dont feel low at all (normally i can tell around 4 that im starting to slide). so i re-tested and was 7.0. iv not had any problems with my meter since then and have tested it against other peoples meters that are the same and it has been about right. just seems a little odd though!

Eddy · #11 (**permalink**) 03-26-2008, 10:47 AM

Quote:

Originally Posted by Lizzie G

one time i got a reading of 2.7 (around 47 in U.S units). i thought WEIRD, lowest ever reading and i dont feel low at all (normally i can tell around 4 that im starting to slide). so i re-tested and was 7.0. iv not had any problems with my meter since then and have tested it against other peoples meters that are the same and it has been about right. just seems a little odd though!

Heh. I had the exact opposite happen a couple days ago. I *knew* I was hypo, and guessed in the low 50s to high 40s. Checked... 147. Nope, not buying it. Checked again... 46.

If in doubt, consider trusting yourself. If you don't want to that... have a little sugar to be safe.

I once had a batch of 100 bad strips. I tested... off-the-meter low. Retested... same. I _thought_ the meter was wrong, but could not swear to it.[1] I chugged a good amount of juice and soda, then proceeded to determine what reality was.

[1] I'd been on the phone for an hour with a friend. I was very relaxed, which reduces my hypo sensitivity.

Glucoweb · #12 (**permalink**) 03-26-2008, 01:37 PM

Does anyone know of a campaign or lobby for the companies to improve their product accuracy? I usually include a paragraph about it in my annual letters to my representatives seeking diabetes funding. But do not know of a specific push for it.

Kim_in_TN · #13 (**permalink**) 03-26-2008, 01:46 PM

I only test once in approximately the same location each time. I can't afford to be frustrated with my readings. I take my first reading and go with that!

JediSkipdogg · #14 (**permalink**) 03-26-2008, 04:39 PM

Quote:

Originally Posted by Glucoweb

I do not believe the quidelines discuss taking the median of multiple readings then make sure all the readings fall within 15% of the median. I did find some interesting information here:
Diabetes Information - Glucose Meters and Diabetes Management
I like the 1993 recommendation, too bad it was never followed and that was 15 years ago.

The 15% is from lab results. So if your lab blood reading was 200, then your meter is perfectly within error from 170-230. Big gap huh? Yeah, it sucks.

Quote:

Originally Posted by Glucoweb

Does anyone know of a campaign or lobby for the companies to improve their product accuracy? I usually include a paragraph about it in my annual letters to my representatives seeking diabetes funding. But do not know of a specific push for it.

There was a while ago but meter companies claim that it's impossible to keep the meters affordable then and a larger sample of blood would be required. You want a very accurate meter, check out the Hemocue Glucose 201. The meter runs $400, uses 5 uL of blood (that's 17 times the Flash and 5 times the Ultra line) and strips are $135 for 100 not covered by any insurance that I'm aware of.

That's why they don't push for more accurate meters, it's just unreasonable in terms of blood sample required and price.