SD Glucometer - accuracy

There's an accuracy range on most meters of + or minus 20%. I'd get one meter and stick with it. A lot of people in the UK seem to use the Codefree, the strips are cheaper, and I haven't really heard any complaints about it.

I bought my Freestyle Lite based on its accuracy and this article.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3570840/figure/fig02/

The trouble is, just because your Accu-Chek is consistently lower than another meter, doesn't mean that it is being accurate - or more accurate than another meter such as the Codefree.

I have tried a number of meters over the years, including the Freestyle Lite, the Codefree and the Tee2.
Also the Freestyle Libre.

And I have found that they each test (fairly) consistently differently.

The Freestyle Lite used to give me lovely low readings, and I was delighted with it.
The Codefree always tested at least 1mmol/l higher than the Lite, and I distrusted it.
But which was right?
No way of knowing.

Then I switched to the Tee2 and it tested slap bang between the Lite and the Codefree, and I decided that I would just stick with it as a rough indication.

And as for the Libre - that always tests about 1mmol/l lower than the Tee2 and suggests that I am heading for a fab HbA1c, yet when the HbA1c arrives, is isn't as low as the prediction.

Honestly - I don't trust any of them to be accurate (including the HbA1c). They are just consistently accurate-ish.
So I just take them as rough ballpark indications, and get on with my life.

If I were a T1 on insulin, I think it would make my brain explode, but I'm not, so I have the luxury of just following the trends.

I agree, as a type 2, that the idea is to look for trends rather than fret over individual readings. I used to use an SD Code Free and in the beginning I was happy with it, however I then had a tub of strips that were reading a lot higher than the others had. I contacted Home Health and they supplied me with a few different batch numbers to compare and I was disappointed with the differences between batch numbers. I lost faith in it I’m afraid. I now use a Caresens Dual which is similar to the Tee2 but has the advantage of being able to measure ketones aswell.

Contour Next from Ascensia has an accuracy reported by the company of +/- 8.4%.

"
The Diabetes Technology Society (DTS) recently revealed long-awaited results from its Blood Glucose Monitor System (BGMS) Surveillance Program. The rigorous study tested the accuracy of 18 popular blood glucose meters (BGM) used in the US. These FDA-cleared meters were purchased through retail outlets and tested rigorously at three study sites in over 1,000 people (including 840 people with diabetes). The results were troubling: only six out of the 18 devices met the DTS passing standard for meter accuracy – within 15% or 15 mg/dl of the laboratory value in over 95% of trials.
The devices that passed were:

Contour Next from Ascensia (formerly Bayer) – 100%
Accu-Chek Aviva Plus from Roche – 98%
Walmart ReliOn Confirm (Micro) from Arkray – 97%
CVS Advanced from Agamatrix – 97%
FreeStyle Lite from Abbott – 96%
Accu-Chek SmartView from Roche – 95%

The devices that failed were:
Walmart ReliOn Prime from Arkray – 92%
OneTouch Verio from LifeScan – 92%
Prodigy Auto Code from Prodigy – 90%
OneTouch Ultra 2 from LifeScan – 90%
Walmart ReliOn Ultima from Abbott – 89%
Contour Classic from Bayer – 89%
Embrace from Omnis Health – 88%
True Result from HDI/Nipro (Trividia) – 88%
True Track from HDI/Nipro (Trividia) – 81%
Solus V2 from BioSense Medical – 76%
Advocate Redi-Code+ from Diabetic Supply of Suncoast – 76%
Gmate Smart from Philosys – 71% "

https://diatribe.org/are-blood-glucose-meters-accurate-new-data-18-meters

There is a paper about this:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5588599/

The SD Codefree tend to measure slightly high but was still very accurate. Bear in mind that measurements from the capillary can never fully replicate venous measurements.

We conclude that the glucometer (SD Codefree) is as accurate as the auto analyser and therefore can be conveniently used as a rapid easy-to-use alternative. However, it is recommended that further studies be carried out using multiple glucometer brands.

Hi ..
Even I am from India and I am using code free for almost one year now. After reading the thread now I am also confused. I taught that the code free wasn’t accurate .. I recently got my HBA1c as 7.0 from lab but my code free use to show always readings around 200 to 200+
So may be my next HBA1c can be an eye opener for code free as my recent readings are shown less than 200 ..

db270 said:

Taught by who? And if you were taught that it's not accurate, then why do you continue using it?

Click to expand...

I will check my HBA1c in oct in lab and than can be sure if code free shows the correct reading for me..

db270 said:

Taught by who? And if you were taught that it's not accurate, then why do you continue using it?

Click to expand...

I taught I saw a puddy cat.

The poster @Reemap meant thought...

Tipetoo said:

I taught I saw a puddy cat.

The poster @Reemap meant thought...

Click to expand...

Yes you are right

db270 said:

Hmm 20% is quite a high margin of error!

Click to expand...

Not for blood glucose meters, apparently. That's the industry standard.

Reemap said:

I will check my HBA1c in oct in lab and than can be sure if code free shows the correct reading for me..

Click to expand...

You can't really compare an HBA1c reading with a meter reading, since it's an average over 12 weeks or so. Unless you have continuous monitoring, there are some spikes you will miss.

db270 said:

Slightly high is how much higher - is it like 5% higher or 20% higher (Sorry I am unable to understand the notations in the paper). .

Click to expand...

Looking at the paper, it seemed to be about 0.7mmol higher (around 13mg/dL) for higher readings (10mM / 180mg/dL) but around 0.4mM(8) for readings around 5mM/90mg/dL so perhaps 3 or 4% too high..

ert said:

I bought my Freestyle Lite based on its accuracy and this article.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3570840/figure/fig02/

Click to expand...

Nice study from 2012. I wish I could find a recent similar one.

alienskin said:

There is a paper about this:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5588599/

The SD Codefree tend to measure slightly high but was still very accurate. Bear in mind that measurements from the capillary can never fully replicate venous measurements.

We conclude that the glucometer (SD Codefree) is as accurate as the auto analyser and therefore can be conveniently used as a rapid easy-to-use alternative. However, it is recommended that further studies be carried out using multiple glucometer brands.

Click to expand...

Thank you for that @alienskin

I have used the Code free since DX
Mainly because of costs, but also as mentioned despite queries as to its accuracy, I too find that whatever method we use it really is for the trend rather then the one off highs or lows.

Are any of them spot on ..?
Debatable, but for my needs it's sufficient.

So nice to get such a good report from a seemingly unbiased angle.

Cheers.

Brunneria said:

The Freestyle Lite used to give me lovely low readings, and I was delighted with it.
The Codefree always tested at least 1mmol/l higher than the Lite, and I distrusted it.
But which was right?
No way of knowing.

Click to expand...

For a long time, and in his latest Teleseminar at the end of July, Dr Bernstein has been recommending the Freestyle Freedom and Freedom Lite as the most accurate meters. (As a renowned diabetes doctor, he has the opportunity to test a lot of meters, as, he says, reps are forever wanting to sell to him). This with the caveat that he considers no meters are accurate once the bg levels rise too far above normal. I can't comment, as the high cost per strip has always deterred me from buying these strips, though I long for their accuracy. I have long used a TEE2 and a Codefree meter. When I get a reading that I don't like, I test the one against the other. Sometimes they are at variance, but usually they more or less agree. Doing the double test helps me in that it forces me to accept readings that are higher than I want. Sometimes, instead of testing with 2 meters, I test a second time with one. The consonance / variability is about the same.