Inaccurate meters

[fkarno]

This is strictly for gear heads.

I have long been worried about the shocking inaccuracy of the BG meters available to patients. These meters get official approval as long as their readings are within +/- 20% of the true figure. In other words, of you get a reading of 5.0, all you know is that your actual BG is somewhere between 4 and 6.

Like everybody, I have a drawer full of give-away meters of which I regularly use only one. The technicians at my diabetes clinic are brilliant - when I go for my three or four monthly check-up, they do the lab test to find my HbA1c and then they take more blood samples and check my meters against the lab result. The differences are horrendous.

Last week my lab BG was 5.1 and three different meters gave results of 6.4 (an error of +25%), 6.8 (an error of +33%) and 7.4 (an error of +45%). With results such as that there is no mileage in using control solutions because those will happily show a pass as long as the error is within +/- 20%. Speaking of mileage, car speedometers by law have a global error allowance of -0%/+10% (disregarding a small additional positive error within a certain speed range). In other words, they must NEVER under-indicate the car's speed but they may over-indicate the speed by 10% which is half the error allowed to our BG meters.

Does anyone know about reasonably priced meters that are more accurate?

Fred

 

[Rob Mitchell]

I found accu chek activ to be closest to my lab readings


Sent from the Diabetes Forum App

 

[Squire Fulwood]

My Freestyle Freedom Lite proved to very accurate in the lower range when measuring the control solution. In the upper range it read a little fast but not by much.

0.05 solution equates to 2.8. Permitted range 1.6 to 3.3. Actual reading =2.8

0.3 solution equates to 16.7. Permitted range 13.8 to 20.7. Actual reading = 17.3

The is for the Freestyle Freedom Lite and using test strips from a batch higher than 1258621 and using the amended ranges.

 

[Yorksman]

The problem with capillary blood is that the blood taken by finger prick samples are a mix of undetermined proportions of blood from arterioles, venules, capillaries, plus interstitial and intracellular fluids. There is no point in making meters super accurate because the method of sampling is only ever going to be a rough guide. Venous blood samples are much better.

A rough guide however is still good, it will provide insights into the effect that certain foods have on you but as an indication of your overall blood glucose management, averages over periods of time should be compared, not individual readings.

 

[parameswaran]

This is strictly for gear heads.

I have long been worried about the shocking inaccuracy of the BG meters available to patients. These meters get official approval as long as their readings are within +/- 20% of the true figure. In other words, of you get a reading of 5.0, all you know is that your actual BG is somewhere between 4 and 6.

Like everybody, I have a drawer full of give-away meters of which I regularly use only one. The technicians at my diabetes clinic are brilliant - when I go for my three or four monthly check-up, they do the lab test to find my HbA1c and then they take more blood samples and check my meters against the lab result. The differences are horrendous.

Last week my lab BG was 5.1 and three different meters gave results of 6.4 (an error of +25%), 6.8 (an error of +33%) and 7.4 (an error of +45%). With results such as that there is no mileage in using control solutions because those will happily show a pass as long as the error is within +/- 20%. Speaking of mileage, car speedometers by law have a global error allowance of -0%/+10% (disregarding a small additional positive error within a certain speed range). In other words, they must NEVER under-indicate the car's speed but they may over-indicate the speed by 10% which is half the error allowed to our BG meters.

Does anyone know about reasonably priced meters that are more accurate?

Fred

I was using AccuCheck active for many years and changed my meter to Easy-Prik basic (I think this is not a well known meter) for last few years. I used to check the lab reading with my meters. Both of them gave reasonably accurate reading (less than 5% variation) in all cases. Hope you are following the instructions properly (such as cleaning the fingers, discarding the first drop of blood etc...).

 

[michaeldavid]

I test my blood sugar around once per waking hour, on average.

But most of the time I don't use a meter. Rather, I mostly use visually read strips: either Betachek Visual or Glucoflex-R. (http://www.betachek.com/uk/)

My blood sugar tends to be exceptionally well controlled. And I find that the visually read strips are essential for the maintenance of that control. (What I eat is crucial too - especially the rye bread.) For not only are they simple, and cheap, but for near-normal readings they are as accurate as one could want. (For anything that looks over 8mmol/l, I'll check again using a meter.)

For readings below 4mmol/l, I find they couldn't really be more accurate.

Also when the reading is low, I find that I can trust the result given a mere 15 seconds after I wipe off the blood. (One must leave the blood on for 30 seconds.)

The visually read strips, moreover, seem to have an inherent failsafe factor. For I have noticed that if my fingertip is not clean, the result on the strip will be a bit smeary in appearance.

Conversly, the numerical result given by a meter (with seeming decimal-point accuracy) only ever appears to be genuine and trustworthy - whether it's actually veridical or not.

 

[jack412]

I was using AccuCheck active for many years and changed my meter to Easy-Prik basic (I think this is not a well known meter) for last few years. I used to check the lab reading with my meters. Both of them gave reasonably accurate reading (less than 5% variation) in all cases. Hope you are following the instructions properly (such as cleaning the fingers, discarding the first drop of blood etc...).

As a Noob this was the first time I have heard about 2nd drop testing. So I googled it and found this interesting.
It seems an unwashed hand should use the second drop.
A washed hand has accuracy for both first and second drop, so it' seems a second drop is unneeded
http://care.diabetesjournals.org/content/34/3/556.full

 

[fkarno]

As a Noob this was the first time I have heard about 2nd drop testing. So I googled it and found this interesting.
It seems an unwashed hand should use the second drop.
A washed hand has accuracy for both first and second drop, so it' seems a second drop is unneeded
http://care.diabetesjournals.org/content/34/3/556.full

Many thanks for this eye opener. It is not always convenient to go and wash my hands before sampling, and I generally have clean paws so I often sample without washing my hands; and this is the first time that I have heard of using the second drop. I will amend my evil ways and see if it makes a difference.

Fred

 

[jack412]

I'm glad you got something from it. Sadly I just lick the side of my finger and wipe it on my pants before I test and use the first drop. Is that close enough?