Variations in Accucheck Performa Nano readings

svgittins

Member
Messages
15
Type of diabetes
Type 2
Treatment type
Tablets (oral)
Some months ago I purchased an Accucheck Performa Nano for testing my BG. Although its not the cheapest device I purchased it as it is made by the medical device division of Roche, the enormous pharma company - and I thought if anyone should be able to produce reliable monitors it would be them. Recently I mislaid the device, and thinking it lost, bought a second one. Shortly afterwards the old one turned up so I now have two meters, both purchased within the past 6 months.

As I had noticed some variations on readings done consecutively on the original device I thought I would do a few tests to see if the results were consistent. This is what I did:
- I obtained a single drop of blood and applied it to testing strips in both of the meters at the same time
- the tests were done sometimes before meals and sometimes 2 hours after eating
- the tests were done over a 3 day period in the past week

The results are below. I see both meters never give the same reading and I was rather surprised the variation was up to 12%. It seems the older meter tends to give higher readings but this is not consistent.

If we had the same type of variation in HbA1c test this could really make quite a difference how we would be treated (e.g. difference between 50 and 45...)

I think my testing methods were reasonable but would welcome suggestions on how I could have been more rigourous. Also am I unreasonable in expecting there to be less variation?

I'd be interested in any comments about this. And does anyone know if all BG meters are subject to the same variance?

I'm Type 2, diagnosed in July 2019. Things well under control now thanks to the hints from forum and continuous monitoring of my BG!

upload_2020-3-13_12-1-29.png
 

Rokaab

Well-Known Member
Messages
2,159
Type of diabetes
Type 1
Treatment type
Pump
I believe all the meters are supposed to be within 15% variance of what your blood sugar actually is, so it is within the allowance variance for them to sell it as working as intended. ie. they never claim to be actually exactly correct.
The HbA1c test is a completely different test so does not have the same variance - but obviously its only an average anyway, and various things can affect it (like anemia I think)
 
  • Like
Reactions: svgittins

porl69

Well-Known Member
Messages
3,647
Type of diabetes
Type 1
Treatment type
Insulin
Dislikes
Stupid people
@Rokaab is correct here. Your meter has a tolerance of 15% so all those reading are well within tolerance. HBA1C is a different test completely and is a 90 day average of your bloods

And now stick to using 1 meter only
 
  • Like
Reactions: svgittins

svgittins

Member
Messages
15
Type of diabetes
Type 2
Treatment type
Tablets (oral)
@Rokaab is correct here. Your meter has a tolerance of 15% so all those reading are well within tolerance. HBA1C is a different test completely and is a 90 day average of your bloods

And now stick to using 1 meter only
Thanks @porl69 and @Rokaab. I had no idea about the 15% tolerance. That would explain why I occasionally have seen post-prandial readings lower than the before-meal ones. I know of course HbA1C is a different test but it did get me thinking if that was always 100% consistent (I have reason to be sceptical of one of my results). Anyhow, different discussion...
 

MarkW3194

Member
Messages
7
Type of diabetes
LADA
Thanks @porl69 and @Rokaab. I had no idea about the 15% tolerance. That would explain why I occasionally have seen post-prandial readings lower than the before-meal ones. I know of course HbA1C is a different test but it did get me thinking if that was always 100% consistent (I have reason to be sceptical of one of my results). Anyhow, different discussion...

HbA1C isn’t a perfect science either, I’ve had a lab run an a1c on blood from 2 tubes of my blood taken at the same time, from the same needle and they were 0.6% different and one taken the next day was over 1% different (I can’t remember the differences in mmol/mol) which is fairly significant.

It’s like anything blood related, the assay has a margin of error and a confidence level, and it’s also why small changes in a result shouldn’t be taken as gospel!