Test BG Meter Accuracy

[Shotokan]

Alright, everybody get your meters out and let's do some tests.

First, for each meter you have use one sample of blood from a finger to get a reading from each meter for that blood sample. Line your meters up and put a droplet on each test strip, then record the results.

Second, repeat the process after one or two minutes with another blood sample taken from the same or another finger.

If you only have one meter, then just do two tests on the same meter one or two minutes apart.

My BG was pretty high when I did the test. Here are my results:

Meter........................First sample...Second sample...Difference...Average
One-Touch UltraSmart............226..................248.... ..........22.......237
Freestyle Flash......................256..................23 0..............26.......243
Accu-Chek Compact...............239..................236.... ...........3........237.5

[lgvincent]

I tend to check my meter against lab results rather than each other.

[sbuff28@charter]

I've never done all 3 of my meters before so i was curious.

Freestyle Flash - 130 123
One Touch - 124 118
BD Logic Medtronic - 127 107


The test in the first colum is from my left middle finger and the second is from my thumb....mabye thats why all my tests were lower the 2nd time.

[psilocybin]

Alright, everybody get your meters out and let's do some tests.

First, for each meter you have use one sample of blood from a finger to get a reading from each meter for that blood sample. Line your meters up and put a droplet on each test strip, then record the results.

Second, repeat the process after one or two minutes with another blood sample taken from the same or another finger.

If you only have one meter, then just do two tests on the same meter one or two minutes apart.

My BG was pretty high when I did the test. Here are my results:

Meter........................First sample...Second sample...Difference...Average
One-Touch UltraSmart............226..................248.... ..........22.......237
Freestyle Flash......................256..................23 0..............26.......243
Accu-Chek Compact...............239..................236.... ...........3........237.5

so judging by your results the accu chek was the most accurate for you

[Shotokan]

Yes, the Compact was most consistent, but it's a very small sample. I can't really hang my hat on it. That's why I'm trying to get more people to do it..increase the sample size.

It's true that testing against lab values is the gold standard, but we'd have to wait weeks or months for people to respond to that.

And you can't really rely on a single comparison of your meter to one lab value. You would know the error for that one reading but you would not know the average error unless you repeated it many times. As you can see, there is a fair amount of variation, even with the same meter, so on a single reading you could get very lucky or very unlucky comparing it to the lab value.

[Dewey]

Alright, everybody get your meters out and let's do some tests.

First, for each meter you have use one sample of blood from a finger to get a reading from each meter for that blood sample. Line your meters up and put a droplet on each test strip, then record the results.

Second, repeat the process after one or two minutes with another blood sample taken from the same or another finger.

If you only have one meter, then just do two tests on the same meter one or two minutes apart.

My BG was pretty high when I did the test. Here are my results:

Meter........................First sample...Second sample...Difference...Average
One-Touch UltraSmart............226..................248.... ..........22.......237
Freestyle Flash......................256..................23 0..............26.......243
Accu-Chek Compact...............239..................236.... ...........3........237.5
I moved your thread to Monitoring, so that more people could see (& maybe partake in the testing of their meters) it & respond. I'll definitely try mine out later on. I have the One Touch UltraSmart, Freestyle Flash & Reli-On Ultima meters that I can try out. Are there specific meters you were looking to compare, or does anything go?

[Shotokan]

Thanks Dewey, I didn't know about this thread. I think we should look at all meters. We will get more results from some meters than others, but we should be able to learn something about all of them.

[condensr]

Onetouch Ultrasmart: 153 129 135
Freestyle Flash: 157 165

[sbuff28@charter]

You would need hundreds of samples to conclude anything significant about the accuracy of the specific kinds of meters. I think right now all it only shows is how the meters can vary drasticly.... almost 20%, The One touch even shows on the test strip bottle that its between 112 and 150 for the control... which is actually pretty bad! Even if it is a Normal distribution....

[sbuff28@charter]

On the same topic as before...


My professor for my class Im taking Applied engineering statistics, makes a firm point to show us exactly how data can be squewed and influenced by whoever toook the data. Believe me that the Qualitry Control Manager at these companies that put out this data ARE misleading you in the data. They'll use confidence intervals and factor in legal obligations and exactly how much they can get away with to mislead you and get you to buy their product. Its all about maximizing profits for all large companies. especially when it comes to advertising.


So it would be really good if we could get the information for ourselves that is unbiased

[lelggren]

My results are as follows: Flash=> 1. 298 2. 281
Ultrasmart => 1. 284 2. 263

My bg's were high when I did these because smart kid me forgot to take my metformin this morning, but my question is, which meter is giving the right reading? The Ultrasmart's reading is certainly more pleasing to the eye, but is it right??? Hmmmm......:hmmmm2:

[spike]

Alright, everybody get your meters out and let's do some tests.

First, for each meter you have use one sample of blood from a finger to get a reading from each meter for that blood sample. Line your meters up and put a droplet on each test strip, then record the results.

Second, repeat the process after one or two minutes with another blood sample taken from the same or another finger.

If you only have one meter, then just do two tests on the same meter one or two minutes apart.

My BG was pretty high when I did the test. Here are my results:

Meter........................First sample...Second sample...Difference...Average
One-Touch UltraSmart............226..................248.... ..........22.......237
Freestyle Flash......................256..................23 0..............26.......243
Accu-Chek Compact...............239..................236.... ...........3........237.5

It's no wonder meter companies stress that one can't compare meter readings from one brand or model to the next. Expecting different meters to read the same at various bg levels is an exercise in utter futility. Just when you think 2 meters agree, you'll run across a disparity that makes you wonder "which meter should I believe right now??". I've gotten myself in trouble more than once by checking my bg with one meter and then picking up a different meter a while later to find a drastic change in the readings. Acting on those numbers has caused me to inject a corrective bolus when I shouldn't have.

I think we should all get familar with the meter we want to rely on, and use that one exclusively (I'm one to talk! I still use several meters!).

Even by using the same drop of blood, I'll get a variation from one test to the next on any meter I've tested, so how can we expect different meters to agree? they don't even agree with themselves!!!

[pbmax]

was at the doc on Monday 3rd april 06 and they use the Accucheck and I use the ultrasmart..
Blood from the same stick pruduced a 90mg on the Accucheck and 30sec later 88mg on the ultrasmart

[seacomp]

a 90mg on the Accucheck and 30sec later 88mg on the ultrasmart
That's as near identical as you'd ever expect to see. And Darn good numbers!:smile:

[spike]

That's as near identical as you'd ever expect to see. And Darn good numbers!:smile:

Don't expect such close numbers when your blood sugar is in the 200-300 range! They might differ by more than 50 points.

[jenet]

Don't expect such close numbers when your blood sugar is in the 200-300 range! They might differ by more than 50 points.
I'm not sure that's always true. :dontknow:

When I did the 3-day CGMS, I had to "mark" points during the day when I tested my BGs using my meter. At the end of the three days, the reports showed my Ultrasmart was within 5-6 pts of the CGMS the whole time, including highs of 350-400 and lows of 45-60. Even my CDE commented on it. When I compare against 3 month lab tests, I also get the same type of accuracy.

Maybe it's just the luck of the draw, but my Ultrasmart seems to be really accurate. If it ever dies, I will be soooo unhappy. :bawling: I'm knocking on wood that doesn't happen any time soon.

cheers,
j

[spike]

I'm not sure that's always true. :dontknow:

When I did the 3-day CGMS, I had to "mark" points during the day when I tested my BGs using my meter. At the end of the three days, the reports showed my Ultrasmart was within 5-6 pts of the CGMS the whole time, including highs of 350-400 and lows of 45-60. Even my CDE commented on it. When I compare against 3 month lab tests, I also get the same type of accuracy.

Maybe it's just the luck of the draw, but my Ultrasmart seems to be really accurate. If it ever dies, I will be soooo unhappy. :bawling: I'm knocking on wood that doesn't happen any time soon.

cheers,
j


What I'm saying is that a Flash, a Compact, and an UltraSmart won't give nearly identlcal readings in the 200-300 range. I do a LOT of testing!

[jenet]

What I'm saying is that a Flash, a Compact, and an UltraSmart won't give nearly identlcal readings in the 200-300 range. I do a LOT of testing!Oooh! I misunderstood. Thanks for clearing up my confusion.
cheers,
j

[spike]

Oooh! I misunderstood. Thanks for clearing up my confusion.
cheers,
j
:) sometimes my posts may not be too clear. sorry about that!

[Shotokan]

I guess it's time to call a halt to this. Thanks everyone who contributed to this thread.

SBUFF28 is right that it would take a large number of readings to get precise estimates of differences among meters and even to get estimates of precision for a single meter. We really need a designed experiment to do a proper statistical analysis.

It's ironic that those of us on pumps can dial in our insulin dosage to the nearest .05 units or .025 units, but our corrections are based on meter readings that are sometimes pretty inaccurate no matter which meter we use.

Some have said that the readings seem to be less accurate as our BG level becomes less normal. I noticed that for my BG readings around 250, the average of the two readings was about the same for all three meters, so maybe we should take the average of two readings when it's really important to get it right, like just before bedtime.

Jenet, when you did the 3-day CGMS trial, you had to calibrate the CGMS to your meter at least 3 times a day, so I suspect that the CGMS readings were close to your meter readings for that reason. That is, the CGMS system "corrected" itself based on the input from your meter. I could be wrong.

Personally, I'm looking forward to the release of the HemoCue meter, which David Mendosa describes here:
http://www.diabeteshealth.com/read,3001,4515.html

Thanks again, everyone!

[jenet]

Jenet, when you did the 3-day CGMS trial, you had to calibrate the CGMS to your meter at least 3 times a day, so I suspect that the CGMS readings were close to your meter readings for that reason. That is, the CGMS system "corrected" itself based on the input from your meter. I could be wrong.
Yes, I understood the CGMS was calibrated with meter BGs, but the CDE said she was surprised by how much more closely they matched than she had seen previously, on a continuous basis for the whole three days. That by itself would only be a little odd and surprising. Having only done a CGMS once myself and not having seen anyone else's, I wouldn't have a clue. :dontknow: The stronger point was that I have similarly accurate results matching my BGs from lab tests (unfortunately also variously high and low).

cheers,
j

[Ronin]

Shotokan, et al.,

This is a very important question. I recently had an e-mail discussion with a fellow tandem bicylist who's day job is an electrical engineer who has designed BG meters. It isn't just the meter, it's also the strips. The strips contain some chemical that works with the blood to allow the meter to measure the electrical current through the sample and thereby determine the amount of glucose in the sample.

While the test Shotokan did says a lot about his meters, a real test of this technology would require a laboratory that would use a test solution with known values, and one test to the next would have to be quite close to prove absolute accuracy. Of course, the laboratory would have to have an incentive ($$$) to do the testing without the incentive coming from the maufacuturer.

Bottom line is that we are dealing with a range of accuracy. It isn't just the meter, it is the strip, the battery, the humidity, phase of the moon, and who knows what else. Then agian, it beats all of those color strips that my father had to use in days gone by.

Personally I'd love to find out how the professional labs come up with their results. Then again, they probably run inside a range as well. As one signature block puts it "this is as much an art as it is a science."

[seacomp]

a real test of this technology would require a laboratory that would use a test solution with known values, and one test to the next would have to be quite close to prove absolute accuracy. Of course, the laboratory would have to have an incentive ($$$) to do the testing without the incentive coming from the maufacuturer.
Given how popular our disease has become, you'd think the Consumer Reports would do a study!
Anyone want to lobby them?
Or how about the FDA? or NIH?

[spike]

Given how popular our disease has become, you'd think the Consumer Reports would do a study!
Anyone want to lobby them?
Or how about the FDA? or NIH?

Good point. Has CR ever reviewed a medical device?

[Tokyo Cate]

Medisense Precision Extra 10.6
Japanese Meter that I pulled out of my earthquake kit (GT-1640) 10.3

Not that different (and not enough to have a huge affect on my decisions).

[sbuff28@charter]

To tell you the truth thats the one thing that scared me the first time i used my meter.I was newly diagnosed in the hospital and My RN was like "Here is your kit, now lets practice..blah blah" and the first test i was somewhere in the 320s, then i did another test i was 256! i remmember washing my hands drying them well and all that and still that really shocked me.did another and it was 293 or something. I spent the next couple days probing doctocs with questions on really how accuate the meters are. Mostly how it really depends on how well you do it, and on what section of the arm or fingers i guess. Still i don't really trust my meters all that much. And I have it much easier than the most of you! I am honeymooning 4+ months and not bolusing. All the sudden i actually feel a lil lucky for once..

[arath13]

Meters are one thing and strips are another. THat is why you need to look at Glucose Oxidase strips which are affected by oxygen(One Touch) and Glucose Hydrogenase strips(precision, accu chek) not affected by Oxygen and then there are the mediators on the strip that need to be looked at. Shall I continue?!?!?!?

[Shotokan]

Please do continue. What are the practical ramifications of the differences?