• Guest - w'd love to know what you think about the forum! Take the 2025 Survey »

How accurate is your meter?

G

Geoff

Well-Known Member
Messages
90
Location
Chepstow, Monmouthshire
Hi all, I went for my three monthly HbA1c test today at my GP's clinic. Nothing unusual about that! But as the nurse was about to draw blood from my arm, I asked her if I could have a bit of venous blood to test with my meter, (I use a Onetouch UltraSmart) as this is the most accurate method for measuring blood glucose, and I wanted to compare it against my usual capillary sample I take by pricking my finger.

My meter read 5.2mmol/lt for the venous sample, I then did the normal finger prick test and got a reading of 4.9mmol/lt, this was quite a surprise as if you read the manual that came with your meter,you will find that a manufacturer can be +- 15-20% variance, and still be considered accurate.

Has anybody else tried matching a venous sample against the normal capillary method with their meter, and what was the result?
 
We have two meters one for school and one for home. On several occations I have tested blood on both meters with same blood sample, and have got different readings.

Both are one touch ultra.
 
Hey Richard, nothing wrong with blubbling man, I turned the waterworks on myself this week for the first time since finding out I was Diabetic :D

Great work on the figures too, you must be well chuffed.
 
Hey, when I first started out it was in the golden age of urinalysis. That's golden as in shower :wink:
The only test available was to pee in a test tube, add some water, drop in a chemical tablet and wait to see which colour it went. Yellow for sugar-free, or blue for full fat coca-cola. Now that was a sophisticated test.
Fancy meters? You lot don't know you're born!

All the best,

fergus
 
Thanks to everyone for giving a reply to this thread, but no one has answered my question which was

Has anybody else tried matching a venous sample against the normal capillary method with their meter, and what was the result?
Click to expand...

Does no one know what I am talking about, or is it that no one has read the actual question?
 
Geoff I haven't ever done that but I definatly will next time I go for a blood test.

Just a question though should the results be the same? I do get confused by meters measuring different thing, some measure blood, some plasma etc,
 
Hi Geoff,
Like (I suspect) nearly everyone else on the forum, I haven't tried this because (a) I don't have the facilities to draw a phial of blood at home and (b) I don't take my monitor when I go for a blood test.

However, logic says that, as the blood that they take from your arm and the blood that you get from a finger prick are both venous blood as opposed to arterial blood, then a test on each should give similar results. The difference that you experienced with your blood test experiment is only a 6% variance, which is well within the manufacturer's tolerance range of >+/- 20%.
 
sofaraway said:
I do get confused by meters measuring different thing, some measure blood, some plasma etc,
Click to expand...
Hi Sofaraway,
I agree it can be confusing when some monitor makers describe the action as plasma testing and other as blood testing. Blood consists of blood cells and platelets that are suspended in plasma. It is the plasma that contains glucose, not the cells, so anything that tests for the presence of glucose is actually testing the plasma content (regardless of what it says on the tin!!).
 
I do get confused by meters measuring different thing, some measure blood, some plasma etc,
Click to expand...

I think the difference that you're confused about is the difference in how the results are reported on your monitor.
Until recently all machines used to report the results as glucose:whole blood but labs reported them as glucose :plasma. The manufacturers in the US then changed to meters giving a glucose:plasma reading. In Europe some now give the results as 'whole boood' some as 'plasma' I think it actually depends on the strips (an accucheck compact in the US gives results as plasma but as whole blood in the UK)
If your machine gives 'whole blood' values they will be about 12% higher than monitors that report 'plasma' values or a laboratory test.

What Geoff was looking at the difference between venous blood and capilliary blood tested at the same time (with the same meter, testing the consistency of the meter).
According to Gretchen Becker there will not be much difference between samples of fasting BS so if your meter reports a similar result then your meter is giving fairly consistent readings, but that after a meal there could be a difference of up to 3.9mmol btween venous and capilliary samples so you wouldn't expect them to be the same.
 
Hi all, sorry for not replying to your questions, I have been out all day at a Time Team dig.

what Nelle said is spot on. The test you do by pricking your finger is a capillary reading, this is not accurate enough to be used for diagnosis of diabetes. A true plasma blood test done by drawing blood from a vain, usually found in the antecubital fossa, this is the area inside the bend of the elbow, must be done for that.

The debate of whole blood v plasma blood readings with your meter is correct, except to add to the confusion, plasma meters do not read true plasma values, but are converted from whole blood to plasma readings by the meters software.

This is why telling someone who thinks they may be diabetic, to do a finger prick test is of little value, unless the reading is way over the normal non-diabetic scale.
 
fergus said:
Hey, when I first started out it was in the golden age of urinalysis. That's golden as in shower :wink:
The only test available was to pee in a test tube, add some water, drop in a chemical tablet and wait to see which colour it went. Yellow for sugar-free, or blue for full fat coca-cola. Now that was a sophisticated test.
Fancy meters? You lot don't know you're born!

All the best,

fergus
Click to expand...

I remember being strongly advised not to stay in the blue too much (chance would be a fine thing) to avoid hypos. NB blue just meant no sugar in the urine, so could be anything below about 9! (There were unofficial shades of blue, paler meant lower, but as far as medics were concerned blue was virtually hypo.) No wonder so many people got complications!

When Alka Selzer ran the 'plink plink fizz' adverts, supposed to be 2 hangover tablets landing in a glass of water, I thought of it as plink (drop of urine), plink (drop of water), fizz (add Clinitest tablet).

Ahh for the fizz of a Clinitest tablet :lol:

Sue
 
Along the same lines as this, I just bought a couple of Accu-Check Compact's as they're cheap right now.

Just set one up, did all the control solution stuff etc and tested myself at 7.5 (just had tea), then I went back to my Ultrascan and that reckoned I was 8.

Is it common to see such a discrepancy between different makes of device?
 
Hi Governor, Yes it is normal, each meter is going to have slightly different software to convert the reading to a plasma blood reading. That is why it is aloud for each manufacture to have up to a 20% difference in reading, but still classified as accurate.

This is why labs only use blood from a vain to get an accurate reading.
 
There are lots of other reasons why finger-prick monitoring isn't hugely accurate. Tiny contaminations of the skin can make a big difference. We all know (I hope) to wash hands before sampling, but how scrupulous is your scrubbing? and how clean is your towel? I was once surprised by a reading of over 10, after which I washed my hands thoroughly for a second time, and dried them on a clean towel to then get a reading of just over 7. Also, meters get bashed around a lot in comparison with lab equipment - which is rarely, if ever, even moved. In a lab consumables are usually kept in a controlled environment (temperature, humidity etc.) and almost never opened until they are needed. Our strips are pulled out of tubes - that may have been open for a while, and are subject to the vagaries of life and climate. The list goes on - it is really quite a testament to the companies behind monitors that they provide results as good as they do. However, you really shouldn't regard them as much more than an estimate - albeit quite a good one. Fortunately, they are generally more accurate at the low end of the scale than the high end, so they are likely to reliably warn you if BG is going dangerously low.
 
Back
Top