My Meter Readings

[Bellx15]

I forgot to ask about this -

My Accu Chek Compact Plus meter has not been checked against a control solution (just ordered one today), but when I recently went to the hospital to give my blood sample, I tested my BG level at home before and after the visit. My meter gave readings of 6.7 and 6.3 taken less than one hour apart. The hospital sample taken in between these two produced a reading of 5.3!

1. Can I infer from this that my meter is reading about 1mmol/l too high?

2. Will the control solution provide a more accurate measure of the machine's error margin?

Thanks.

 

[Squire Fulwood]

I forgot to ask about this -

My Accu Chek Compact Plus meter has not been checked against a control solution (just ordered one today), but when I recently went to the hospital to give my blood sample, I tested my BG level at home before and after the visit. My meter gave readings of 6.7 and 6.3 taken less than one hour apart. The hospital sample taken in between these two produced a reading of 5.3!

1. Can I infer from this that my meter is reading about 1mmol/l too high?

2. Will the control solution provide a more accurate measure of the machine's error margin?

Thanks.

This sort of thing has been discussed on the forum and I now accept that there is no such thing as as "accurate" meter. Zurich University did a much quoted study of them and no two makes of meter read the same so where is the accuracy in that. Some of them are famous for reading high. The SD codefree is one in point.

You cannot infer that your meter is reading 1mmol/l too high without much comparison with an accurate lab measurement but it probably is so. All I have done on the subject is some 50 readings where I compared a SD codefree to a Freestyle Lite and determined that the SD did read 1unit higher than the Freestyle. I only did that so that I could talk to my nurse in figures that she was used to .....the Freestyle.

I doubt if the control solution will help you much since they usually only give a very wide range of possible readings and if your meter is anywhere inside that range then it is ok ..........they say.

You will have to get over the accuracy quest. No home meter is guaranteed to be accurate or anything close. If your readings are persistently one higher than the doctors surgery then just note that somewhere and if you wish to knock one point off your readings then make sure you know what you are doing.

 

[Bellx15]

Thanks!

The way I understand it (or don't) is that the control solution is set at an accurate level of, say, 8.5, and the manufacturers are just saying any reading between 6.9 and 9.1 is within their specified level of accuracy. If I am getting that right, then, I would know exactly what the glucose content of the solution is and be able to see from my meter reading exactly how accurate it is.

Is that about right?

 

[Squire Fulwood]

Thanks!

The way I understand it (or don't) is that the control solution is set at an accurate level of, say, 8.5, and the manufacturers are just saying any reading between 6.9 and 9.1 is within their specified level of accuracy. If I am getting that right, then, I would know exactly what the glucose content of the solution is and be able to see from my meter reading exactly how accurate it is.

Is that about right?

I don't know how your manufacturer performs but I have just been all over the control solution bottles and test strip containers of my Freestyle with a magnifying glass and it does not say what sugar content the control solution is.

The test strips give ranges where they are deemed to be ok if the reading falls between the limits of the range. I could find the exact centre of that range and make assumptions that that is where the manufacturer suggests the actual sugar content is but I would be making it up.

My control solutions have one with a blue cap meaning low and one with a red cap meaning high and no mention of how low or how high,.

 

[Bellx15]

OK - I'll ask the manufacturer then. Only way.

 

[Squire Fulwood]

OK - I'll ask the manufacturer then. Only way.

Great idea. Let me know if they can give you a figure.

In the meantime I got my old test strip pots out and found five from different batches. They all give the same ranges so this could be based around a particular test solution strength. Your manufacturer will give us a clue.

 

[Bellx15]

I will report back.

Actually, though, I do think your interpretation (mid-point between stated tolerance limits) must be about right, as the law requires a certain percentage (+/-) accuracy.

 

[Squire Fulwood]

I will report back.

Actually, though, I do think your interpretation (mid-point between stated tolerance limits) must be about right, as the law requires a certain percentage (+/-) accuracy.

Maybe some of the tolerance is built in for variations in the control solution as well as the test strips. It would be interesting to know what your manufacturer says and later I may contact mine.

 

[Bellx15]

Could well be so.

I just phoned them and the woman seemed to have no idea. I had to explain the question three times, and then she said she's never been asked that before. I find this incredible. What i also find hard to believe is the wide margin they permit in your test results. My strips specify 6.7 - 9.0 mmo0l/l as the limiting readings permitted. That means an accuracy of only +/- 15% (assuming midpoint is the actual glucose content).

She's "going to call me back". :roll:

 

[Squire Fulwood]

I emailed mine because I thought they may need notice of the question.

My test strip limits.

Low 2.2 to 3.9
High 13.8 to 20.7

I wonder if these are not so much the practical limits but more like the legal ones. In any event a variation of the control solution could well be tolerated.

 

[Bellx15]

Yes, they are the legal ones; meters have to give readings less than 15% out more than 90% of the time (or something along those lines).

Problem for me is that my actual readings appear to be at least 1.0 mmol/l above the correct readings, if the hospital is accurate. That's a lot.

 

[Squire Fulwood]

Problem for me is that my actual readings appear to be at least 1.0 mmol/l above the correct readings, if the hospital is accurate. That's a lot.

Yes I understood that but I thought that knocking one point off on the basis of just one comparison might be a little rash. When I did the comparison of my two meters I occasionally got spoof readings which I had to explain or disregard. That's why I did 50 readings with 24 of the tests comparing both readings from the same drop of blood (but a different drop for each test if you know what I mean).

This is proving interesting.

 

[catherinecherub]

I forgot to ask about this -

My Accu Chek Compact Plus meter has not been checked against a control solution (just ordered one today), but when I recently went to the hospital to give my blood sample, I tested my BG level at home before and after the visit. My meter gave readings of 6.7 and 6.3 taken less than one hour apart. The hospital sample taken in between these two produced a reading of 5.3!

1. Can I infer from this that my meter is reading about 1mmol/l too high?

2. Will the control solution provide a more accurate measure of the machine's error margin?

Thanks.

Have you read this thread explaining why meters cannot tell us our BS readings?
viewtopic.php?f=1&t=32762&p=308135&hilit=why+meters+cannot+tell+us#p308135

 

[Squire Fulwood]

Thank you for the link. Informative and makes you want to give up. Well, no not that bad but luckily all I have to do is to produce readings similar to those that my nurse expects.

 

[Bellx15]

The thing is that i took two readings around the hospital one, which I understand is super-accurate.

If I consistently get such high readings with the control solution (e.g., around 8.5 - 9.0) I think it will be fair to say my meter is about 1 mmol/l over.

No phone call.

 

[Squire Fulwood]

The thing is that i took two readings around the hospital one, which I understand is super-accurate.

To add to the variables here is a paragraph from the article that Catherinecherub indicated earlier.

"Cariski says lab tests generally come within about plus/minus 4% of a perfect reading. Andreas Stuhr, Medical Director North America at Roche Diagnostics, more or less confirmed that but added: “Even the lab standard is off 5-7% at times, so plus/minus 20% of the hospital lab test (current ISO standard) isn’t as big a leap as we think.”

 

[Bellx15]

Yes - as this conversation is developing, so is my opinion -

That the only way to do this is by running a sequence of home control solution readings. I think we really have to assume that the control solution falls about mid-way between the specified limits. If your meter reads consistently high or low for the control, then adjust your BG readings by that amount.

Still no phone call, but I too emailed them at the same time.

 

[Squire Fulwood]

There are far too many variables in the mix now for any meaningful result. Your method will also have to cope with the fact that test strips may vary from batch to batch.

I once said that accuracy was a Holy Grail that we could not pursue. Now I am not sure what I mean by accuracy.

Mind you when you are waiting in for Yodel to deliver a parcel it gives you something to do.

 

[Squire Fulwood]

I got a reply from Abbott and they asked for the model of my meter, the serial number of my meter and information about the control solution. I supplied all this information and asked them how that helps them provide me with the intended sugar content of their control solution.

I think this could be a long one.

 

[Bellx15]

Could be entertaining though ...

By the way, I'd appreciate an opinion on this:

I haven't done a proper GTT test but for now I thought I'd monitor my response to a light tea. It consisted of three small slices of wholemeal bread, one fried egg, butter, a substantial blob of peanut butter, two glasses of white wine.

Results (probably higher than true readings because of the machine): Before the food - 6.5, 1 hr after the food- 6.8. 1.5 hr after the food - 7.9, 2hr after the food - 7.4

What do you make of that ??? :***: