Meter Variance (Re: one of Grant's post)

[jer.lawrence]

I've quoted a reply from Grant in another thread below, and he makes a very good point that I sort of have questions about as well.

I guess knowing that my meter can be within that 20% range, I try to pay more attention to the trends of the numbers over a period of time rather than the specific numbers themselves.

However, when you're on insulin, or even when you're trying to determine an average over a period of time, how do you rely on your meter knowing that it can be off by as much as 20%?

Are there certain meters that are more accurate than others? How often have you tested against Lab results, and how close is it?

Just curious, as I'm new getting into this, and will be at it for a very long time, I'm sure.

Thanks!

I was astonished to find out yesterday that my one meter 'when it's under 100' is NOT as accurate. I've done lots of comparisons between my meters when my numbers have been mostly above 100 and the two meters track nicely. But when my ReliOn Micro shows an 80-95 number, my Freedom Lite meter shows 94-115. I've done control solution tests on both -- and they're both in NOMINAL range in the test. So that's not it.

So many here have talked about the 20% variance you can get in meter reading and yet so many of us live our lives as if our meters are 'dead on' with no variance and we make 'choices' and 'determinations' about these numbers. In your mind a 99 would be WAY different than an 85 and yet that's within 20% variance. Do you often check with multiple meters? I'm starting to take both my meters everywhere with me. It's more expensive to test with two -- but I need to know the 'average' when I'm dealing with dosage changes. What a drag. I need a meter that gives me total accuracy. How do the Type 1's do this? They need to get 'pin point' accuracy in determining bolus insulin. I'm frustrated by the difference in my meters under 100 in the reading. I wonder if your numbers are perfectly accurate. Though obviously your pre-diabetes is controlled on your diet, Shottle. Keep up the good work.

[fairyblood]

Honestly I don't worry about the 20% I take my meter at its word unless there is a very good reason to think otherwise. However there are times when I trust my body over my meter. So if I am around 100 and feel like I am dropping I will treat then. I often feel high when I first wake up so if I'm not then I trust my meter and not my body.

There are some meters that run higher then others. For example freestyle tends to run higher then some of the other meters on the market. I happen to like that because I like treating things aggressively and this helps me do that.

The only time I test a second time is if it feels really off. So if I feel low or fine and it comes out high then I will test again. I personally think it is both a waste of money and a bad idea to constantly test on two meters at the same time. I can barely afford my strips as it is. Diabetes is full of inconsistency and sometimes you have to accept that things aren't always right or clear. You can drive yourself batty worrying about whether you are 95 or 85. I personally am more then happy with either number.

As a type one there are so many factors out of my control that I just take my number do what I need to do and move on.

[fgummett]

Bottom line is that apparently many of us with D -- even those on insulin -- can achieve very good BG control despite meter inaccuracy... the situation is so much better than before home BG testing when all that was available was urine testing... which basically showed such a high BG that it was spilling out of the kidneys. It is not an exact science... there are way to many variables to consider, hence the need to take multiple data points over many weeks and months so that you can become comfortable with repeatable patterns, rather than jumping to conclusions after a few observations... I'd suggest sticking with one meter and carrying it with you.

[jps]

Though the claims are 20% inaccuracy, most are much closer than that. They widen the gap probably to protect themselves. I have a One Touch II and I've measured several venous samples from different patients over various points against my Beckman Coulter DxC (heavy duty laboratory analyzer) and my particular meter was about 2% lower on average.

[NewdestinyX]

Thanks, Jer, for starting this thread!There are some meters that run higher then others. For example Freestyle tends to run higher than some of the other meters on the market. I happen to like that because I like treating things aggressively and this helps me do that. Wow.. didn't know that. My one meter IS a FreeStyle Lite and it's high to my ReliOn Ultra.

You can drive yourself batty worrying about whether you are 95 or 85. I personally am more then happy with either number.Yes -- I think that can be true. What gets weird and where I wish I had 'perfect accuracy' under 100 is when you're setting Basal dose which is a process I'm in now with Lantus. 85 or 95 makes a huge difference there.

It is not an exact science... there are way to many variables to consider, hence the need to take multiple data points over many weeks and months so that you can become comfortable with repeatable patterns, rather than jumping to conclusions after a few observations... I don't tend to jump to conclusions after a 'few' observations but I make determinations over weeks. If for weeks I've been averaging 89 but it's really 104 then my conclusions for my control regimen are off. That's the point... I think... :o I'd suggest sticking with one meter and carrying it with you.I think I may move to that system. It's easier to have one at work and one at home. Too bulky to carry around in my experience.

Though the claims are 20% inaccuracy, most are much closer than that. They widen the gap probably to protect themselves. This is my experience with numbers above 100 -- so I thought it was true universally. I have a One Touch II and I've measured several venous samples from different patients over various points against my Beckman Coulter DxC (heavy duty laboratory analyzer) and my particular meter was about 2% lower on average.Lucky dog -- having access to such a piece of equipment... ;) Well since you've already tested the OneTouch II - I may pick that one up. Does it have a port for USB transfer to computer?

[jps]

I'm grateful that I have access to all that testing. I give partial credit to that for gaining control. It's easier to truly determine what is going on when you can run a full lipid panel - or whatever - every single week.

It's extra, but the One Touch II does have a very easy and efficient software program with a data port.

[Grunch]

My breeze 2 meter is way more accurate than 20%. The accu-check on the other hand seems very inaccurate, I don't trust it at all.

[NewdestinyX]

My breeze 2 meter is way more accurate than 20%.
How do you verify that? I ask because in all my searching of late it's the Breeze 2, with its test strip DISCS and USB transfer to computer that most interests me! The strips (divided by 10 in a disc) are only a little more expensive than my Freestyle Lite ones on eBay.

[Grunch]

Just by observation, I can't guarantee it. When I test 2 times in a row I've never seen a difference close to 20% and when I start feeling low it's very often around 65 but has never been above 70.

The disc is the coolest part. I hate individual strips.

[fairyblood]

I have heard that wavesense meters are supposed to make some of the most accurate meters on the market for the consumer.

I have a One Touch II

I am surprised that anyone still uses the One Touch II. That was the meter that was the very popular when I was dx. in 1996. It takes a lot of blood and requires cleaning. Is there a reason that you use that meter or just personal preference?

when you're setting Basal dose which is a process I'm in now with Lantus. 85 or 95 makes a huge difference there.

My life is insulin... and as a type one I am much more likely to go low then a type II. I still don't see how whether it is 85 or 95 could make much difference at all. I can see how emotionally it could feel different but otherwise they are almost the same number.

[jps]

Sorry, One Touch Ultra 2. I was given two meters, both the same. One by the CDE, one by my physician.

[fairyblood]

Sorry, One Touch Ultra 2. I was given two meters, both the same. One by the CDE, one by my physician.

Ahh good. Glad to hear you aren't in the meter dark ages. ;)

The One Touch II really was a beast. I am sure people still use it but I don't know why you would.

[jer.lawrence]

Sorry, One Touch Ultra 2. I was given two meters, both the same. One by the CDE, one by my physician.

I'm also glad to hear that because I use a One Touch Ultra 2.

I got a scary reading after dinner tonight of 149 (I didn't THINK I had too many carbs, but ate out and wasn't sure). I washed my hands and re-tested at 103. Still not satisfied, I tested one more time and came back at 101. I think that first test must have just been tainted with something on my hands.

I'll probably do a check on myself next time when I get my blood work done as well to try to gauge how accurate it is.

[jtausch]

I did that twice on my acue-check aviva meter once with a meter that the local hospital used it was within-2, the second time at the lab with a sample taken from the blod vial it was -1 one off the the lab test I think my meer is accurate. I think a lot of it has to do with the strips they could also be off to.

The blood sample reacts with the chemicals in the strip and generate a very small voltage and this is what the meter reads and turns it into a number.

[fgummett]

I have also noted a variation in batches of test strips... one bottle may tend to read higher or lower than the previous. As mentioned above, it was recommended to me that I take my meter along when having a routine lab Fasting BG and compare the readings -- I test on the hand of the same arm as the blood test within 5 minutes of the sample being taken.

[Subby]

You can drive yourself batty worrying about whether you are 95 or 85. I personally am more then happy with either number.

As a type one there are so many factors out of my control that I just take my number do what I need to do and move on.

Jeremy, I think you'll find that's the kind of utilitarian approach taken up by many or most type 1s. This apparent need for "pin point accuracy" is simply a pipe dream if it doesn't exist, we have enough to worry about and to get on with what with managing the condition and having a life, so why sweat it if you've got "pretty good"? You do what you can with what you have. Certainly Type 1s can gain great control despite meter inaccuracy. And I'll wager that those of us who do have more difficulty with control, there are much bigger factors, some potentially alien to some type 2s, that may well dwarf meter inaccuracy as a control issue.

That said, I do what I can to keep an eye on my tester over time. Test against lab results. I will take a few tests if I suspect a number is wrong. I have another relatively accurate meter I will use to compare if I am really suspicious. If a meter gives me a lot of inaccurate results, it gets turfed, simple.

If and when affordable testers are significantly more accurate, I will celebrate. I think it will make our life easier and improve insulin therapy a little. Til then, as I said, you do with what you've got.

[notme]

Jeremy, I think you'll find that's the kind of utilitarian approach taken up by many or most type 1s. This apparent need for "pin point accuracy" is simply a pipe dream if it doesn't exist, we have enough to worry about and to get on with what with managing the condition and having a life, so why sweat it if you've got "pretty good"? You do what you can with what you have. Certainly Type 1s can gain great control despite meter inaccuracy. And I'll wager that those of us who do have more difficulty with control, there are much bigger factors, some potentially alien to some type 2s, that may well dwarf meter inaccuracy as a control issue.

That said, I do what I can to keep an eye on my tester over time. Test against lab results. I will take a few tests if I suspect a number is wrong. I have another relatively accurate meter I will use to compare if I am really suspicious. If a meter gives me a lot of inaccurate results, it gets turfed, simple.

If and when affordable testers are significantly more accurate, I will celebrate. I think it will make our life easier and improve insulin therapy a little. Til then, as I said, you do with what you've got.


Couldn't agree more. I use the One Touch that goes with my Medtronic pump. I have tested it against lab work and it appears to be well within acceptable limits for me. My last fasting blood test said I was 109 and my meter gave me a 104 before the test. Good enough for me!

[jps]

I have also noted a variation in batches of test strips... one bottle may tend to read higher or lower than the previous.

There are certain tests in my lab that are notorious for this, it just depends on the nature of the test. My glucose reagent, for example, shows absolutely no variation from lot to lot (batches). Same can be said for about 90% of my reagents. However, there are some that show some lot to lot variability and it drives me absolutely crazy. For example, Vitamin B12 and Folate. I can have a two month supply of Lot "A" and my low control (we run three levels) will yield a mean of 150 with a Standard Deviation of 5 giving us a valid range of 140-160. Then we run out of that lot and go to Lot "B". After a couple weeks of testing, the mean will be revealed as 165 with a Standard Deviation of 8 giving us a new valid range of 149-181. So what that's basically telling me is if we run a B12 on Patient A on the last batch of Lot A and his value is 200, when the doctor orders another B12 on him in 3 days, he can have virtually the same exact levels of B12 in his system, and we will be reporting out a 220. Still well within the limits of laboratory standards, but I despise it. I want less than 5% variability on my tests.

Sorry for the long winded post. But essentially, yes, THERE CAN be variability from lot to lot with your glucose strips. DO NOT be surprised (or panic) if you see a slight "shift" in your values when you switch lots. You may not see it at all - again, its dependent upon the manufacturer and their standards for the test. Overall, they should still be relatively accurate though.

There's no question that manufacturers can minimize this variability and they can certainly increase the precision and accuracy, but that would involve improvement in research and development and much tighter quality control on their end.

What does that mean? A much higher cost to the end user.

[fgummett]

I noticed the test strip batch/lot variability some time ago but just this week noticed that the back of the test strip bottle actually support this as well...

I have 2 bottles of Contour test strips: lot 8MC3B03 shows 6.0-8.3mmol/l as acceptable control solution values, while lot 8HC3C02 lists the range as 5.9-8.2mmol/l

It doesn't phase me anymore... I just note it and move on. If I am extra keen, I will test with the last strip from the old bottle and also with a strip from the new one, just as I switch over.

I did read somewhere about an accurate unit made in Scandinavia (?) for around $600, with test "strips" at around $5 a pop -- that had to be refrigerated before use -- and IIRC required several minutes for a test.

[Tribbles]

The Norwegian health authority get meters independently tested before issue and for three batches of strips quite often one batch will be noticeably different. That is under lab conditions.

[jcaboose]

I'm new to this whole thing, just diagnosed with Type 2 earlier this year. I went to an Endocrinologist today, as I didn't think I was getting much in the way of help or direction from my PCP. When the nurse took my initial info, she did an "instant read" test, and my glucose level was 123. I had never been given, or told to get, my own meter. At the end of my appointment (about 2 hours later), they gave me a Bayer Contour meter and showed me how to use it, and my reading was 95. They also gave me a prescription for 200 lancets and test strips, which I took to my regular CVS to get filled.

It just so happened that they were having some kind of Diabetes Awareness thing today, and there was a guy from Abbott Labs there, and we started talking while I was waiting in line. They were giving away their Freestyle meters, so he gave me a Freestyle Lite and showed me how to use it, and my reading was 125.

I came home and had dinner, waited an hour, and then tested with both meters. The Contour said 153, and the Freestyle said 121. That didn't seem to make a lot of sense, so I figured I must have screwed something up, so I went to do it again and make sure I did it right - I washed and dried my hands again, put test strips in and got both meters ready, and then used a new lancet on a different finger. I used the same drop off blood, and this time the Contour said 141 and the Freestyle said 125.

I realize there's going to be some variance, but can I believe either of these monitors? Which one? The Freestyle seems to be more "consistent," but is it just wrong all the time by about the same amount? Is the Contour just a piece of junk and as reliable as throwing darts at random numbers? Is there another meter/system I should ask for instead?

I don't want to be freaking out if my numbers are jumping all over the place or reading higher than they really are, nor do I want to think "oh look, my numbers are normal, I don't have to worry about this any more" when my glucose is actually too high.

Sorry to prattle on, but any advice on meters and monitoring gratefully accepted.

Just tested again - Contour 187, Freestyle 151. Oy vey.

[Grunch]

There's something I don't understand.

My strips say this on the bottom:

cont L: 39-68
cont N: 89-121
cont H: 224-292

I understand those are measured on some kind of control solution but what do the numbers mean? Is it that the strips might measure anything from 39-68 on the same drop of blood? That can't be correct...

[fgummett]

There's something I don't understand.

My strips say this on the bottom:

cont L: 39-68
cont N: 89-121
cont H: 224-292

I understand those are measured on some kind of control solution but what do the numbers mean? Is it that the strips might measure anything from 39-68 on the same drop of blood? That can't be correct...Perhaps there are three different control solutions: one to simulate Low BG, another for Normal and a third for High. In that case the manufacturers acceptable values for the normal control solution would range from 89-121..?

[jcaboose]

First thing in the morning, no food for 8+ hours:
Contour 191, Freestyle 150. These two monitors are well over 20% apart.

Confused and frustrated.

[NewdestinyX]

There's something I don't understand.

My strips say this on the bottom:

cont L: 39-68
cont N: 89-121
cont H: 224-292

I understand those are measured on some kind of control solution but what do the numbers mean? Is it that the strips might measure anything from 39-68 on the same drop of blood? That can't be correct...Perhaps there are three different control solutions: one to simulate Low BG, another for Normal and a third for High. In that case the manufacturers acceptable values for the normal control solution would range from 89-121..?Only one solution is made per a chat with my company. It's supposed to test in the 'nominal or normal' (N) range. If the solution brings you a number in the 'H' or 'L' ranges - the strips are bad. (But the meter could be bad which is weird too).

[Subby]

First thing in the morning, no food for 8+ hours:
Contour 191, Freestyle 150. These two monitors are well over 20% apart.

Confused and frustrated.

Unfortunately it doesn't work that way, they could both have been within the 20% error range.

191 could have been top of a range approx 155 - 191

150 could have been bottom of a range approx 150 - 180

I say approximately, because causally working out the exact percentiles involved with a theoretical "accurate number" is not really useful, nor do I have the maths acrobatics to do it easily. These are rough illustrations.

That said, I really understand your frustration. I'd suggest testing both meters, multiple times, against lab results - soon if possible. It's worth remembering and doing the sums sometimes to illustrate for yourself - the higher the BG, the bigger the actual territory of error. With lower Bg readings, where I would suggest there is a lot more importance for some accuracy, the 20% gets narrower and narrower. That's something at least.

In the short term, if you notice a big discrepancy between meters or tests results at the same time, I suggest rewashing hands and taking another test to allow for another point of reference.

[fgummett]

Only one solution is made per a chat with my company. It's supposed to test in the 'nominal or normal' (N) range. If the solution brings you a number in the 'H' or 'L' ranges - the strips are bad. (But the meter could be bad which is weird too). So what about numbers between the low end of the Normal range and the top end of the Low range??

[NewdestinyX]

So what about numbers between the low end of the Normal range and the top end of the Low range??:confused: I guess I'd be suspect and ask the company to send me a replacement bottle... I guess...

[fgummett]

Or... perhaps YOUR company only makes one test solution and OTHER companies might make Low, Normal and High control solutions :cool:

[NewdestinyX]

Or... perhaps YOUR company only makes one test solution and OTHER companies might make Low, Normal and High control solutions :cool:Maybe.. would be weird if that were the case since my company lists 3 ranges on the bottle too yet only makes one solution. All you need to determine bad strips is a number 'out' of the N range. That part seems simple to me.

[fgummett]

Maybe.. would be weird if that were the case since my company lists 3 ranges on the bottle too yet only makes one solution. All you need to determine bad strips is a number 'out' of the N range. That part seems simple to me.Talk about always having to have the last word :mad: I agree it is "simple" but your response does not explain the other two listed ranges nor the gaps between the ranges. The fact that a web search and my personal experience turns up Low, High and Normal control solutions DOES.

Try a web search on "Glucose Control Solution High"

[NewdestinyX]

Talk about always having to have the last word :mad: I agree it is "simple" but your response does not explain the other two listed ranges nor the gaps between the ranges. The fact that a web search and my personal experience turns up Low, High and Normal control solutions DOES.

Try a web search on "Glucose Control Solution High"My goal isn't to answer any one person's every question, Frank, but to off my 'experience' - which is what I've been 'encouraged to do more of'. Conjecture's beyond 'my experience' get me into trouble around here... I made the caveat that 'with MY company'. That's my experience. :banghead:

When you keep asking a 'clarifying question' I will always have a response. That's not needing the last word. I'm glad other companies make multiple test solutions. That seems like a 'great idea'.

[jps]

There's something I don't understand.

My strips say this on the bottom:

cont L: 39-68
cont N: 89-121
cont H: 224-292

I understand those are measured on some kind of control solution but what do the numbers mean? Is it that the strips might measure anything from 39-68 on the same drop of blood? That can't be correct...


Typically, the way these things are figured out: They'll take a relatively known low solution. They will run it thousands of times against many different meters of the same make. They'll gather the data and crunch the numbers to come up with the range. For the low control you speak of, the mean of all the numbers of the data group would be 54. One standard deviation should be, at most, 10% of that (5.4). All control values should be within 2 standard deviations of the mean (5.4 x 2 = approximately 11). Subtract (and add) 11 from the mean to get your range = 43-65. ...Obviously, they allow for a little bit more error.