Test results, wide variations

HSSS

Expert
Messages
7,465
Type of diabetes
Type 2
Treatment type
Diet only
i just did 4 tests back to back on the same monitor, an accuchek mobile, from the same pinprick. (Testing a low carb slice of bread 2hr , not great whichever it is)

9 (wasn’t expecting this much so repeated)
7.2 (more expected)
8.7 (3rd sample for confirmation of 2nd)
7.7 (giving up now confused)

What am I supposed to make of that? How can I trust any readings?
 

Bluetit1802

Legend
Messages
25,216
Type of diabetes
Type 2 (in remission!)
Treatment type
Diet only
Meters are only required to be accurate to within plus or minus 15% (and then only 95% of the time). It is swings and roundabouts. Some readings will be higher than true, some will be lower, but at the end they will work themselves out. All we can do is record and watch for trends. The trends will show themselves, with just an odd high or low thrown in the mix.

When I feel the need to repeat test I tend to average them, or see if there is a cluster much the same. I also clean my hands between each test and never use the same finger prick.
 

Alexandra100

Well-Known Member
Messages
3,738
Type of diabetes
Prediabetes
Treatment type
Tablets (oral)
i just did 4 tests back to back on the same monitor, an accuchek mobile, from the same pinprick. (Testing a low carb slice of bread 2hr , not great whichever it is)

9 (wasn’t expecting this much so repeated)
7.2 (more expected)
8.7 (3rd sample for confirmation of 2nd)
7.7 (giving up now confused)

What am I supposed to make of that? How can I trust any readings?
These discrepancies are SO annoying, but at least it seems clear this bread is not the greatest for you. However you really need to know whether your 2 hour number was actually the peak of your bg rise, on the way up, or on the way down. For this you need either a Libre or to test eg at 1 hour, 2 hours and 3 hours or even later.
 

HSSS

Expert
Messages
7,465
Type of diabetes
Type 2
Treatment type
Diet only
I did wash hands. I thought using the same finger prick would keep consistency. For now I’ll go with the average as suggested. So glad I’m not injecting on the basis of these monitors though.

I’d love a Libre at least short term and have been reading about it but seems v expensive for a type 2 on no meds that seems to be managing without. And yes the bread is obviously not good and may have been even worse than this at 1 or 3 hrs.
 
  • Like
Reactions: Serena51

Bluetit1802

Legend
Messages
25,216
Type of diabetes
Type 2 (in remission!)
Treatment type
Diet only
I’d love a Libre at least short term and have been reading about it but seems v expensive for a type 2 on no meds that seems to be managing without.

You could treat yourself. I am T2 on no meds with normal BS levels but I still use a Libre on a part time basis. They are illuminating. I find them especially useful when on holiday.
 
  • Like
Reactions: Energize

Gork

Well-Known Member
Messages
71
Type of diabetes
Type 2
Treatment type
Insulin
I get similar errors with sequential readings. More so, using the freestyle libre system, the accuracy changes sensor to sensor. In 11 sensors, two had 20% error but 9 had less than 8% error. I was comparing the sensor readings with test strip readings taken with the same blood sample.
 

KK123

Well-Known Member
Messages
3,967
Type of diabetes
Type 1
Treatment type
Insulin
I find this SO annoying as well. Like many others I try to keep my levels at the lowest possible without hypos (not easy). I rely totally on a monitor reading so if it's 4.4 for example I leave well alone, however that 4.4 could actually show as 3.4 on another day or 5.4 on yet another. How can I adjust when the readings are a guesstimate at most. Also I could take a reading to try and see whether I was on the way down or up but of course with inaccurate readings (even on the same monitor which I have calibrated with other monitors) that's also a guess. It's like the Grand old Duke of York.
 

Glucobabu

Well-Known Member
Messages
248
Type of diabetes
Type 1
Treatment type
Insulin
Meters are only required to be accurate to within plus or minus 15% (and then only 95% of the time). It is swings and roundabouts. Some readings will be higher than true, some will be lower, but at the end they will work themselves out. All we can do is record and watch for trends. The trends will show themselves, with just an odd high or low thrown in the mix.

When I feel the need to repeat test I tend to average them, or see if there is a cluster much the same. I also clean my hands between each test and never use the same finger prick.
One would expect the technology to evolve and become more accurate in this day and age. It has hardly changed in last 25 years apart from giving us unnecessary Bluetooth and downloadable data to smartphones!
 

Alexandra100

Well-Known Member
Messages
3,738
Type of diabetes
Prediabetes
Treatment type
Tablets (oral)
I find this SO annoying as well. Like many others I try to keep my levels at the lowest possible without hypos (not easy). I rely totally on a monitor reading so if it's 4.4 for example I leave well alone, however that 4.4 could actually show as 3.4 on another day or 5.4 on yet another. How can I adjust when the readings are a guesstimate at most. Also I could take a reading to try and see whether I was on the way down or up but of course with inaccurate readings (even on the same monitor which I have calibrated with other monitors) that's also a guess. It's like the Grand old Duke of York.
Have you tried using an Abbott meter and strips, which Dr Bernstein says is just about the only one that is accurate? I fear the strips are very pricey. I am listening to a podcast of Dr B's latest Teleseminar today, and he says all the other meters are grossly inaccurate. I am hoping that with the continuing development of affordable CGMs the strip sellers will have to clean up their act, as the only reason to buy their stuff will be to check the CGMs against something more accurate.
 

Zilsniggy

Well-Known Member
Messages
428
Type of diabetes
Type 2
Treatment type
Insulin
9 (wasn’t expecting this much so repeated)
7.2 (more expected)
8.7 (3rd sample for confirmation of 2nd)
7.7 (giving up now confused)


These machines all have a 15% tolerance factor. All these readings are within that. Why did you have to test so many times?
By the way, lab testing has similar margins, and you will find slight differences in testing from lab to lab.
Testing things like low carb breads, you may need to keep testing after the 2 hour mark, since often they induce rises which take either longer to peak, or longer to fall, depending on ingredients. I think most of us low carbers eventually realise that we just can't have bread anymore. I don't really miss it
 

HSSS

Expert
Messages
7,465
Type of diabetes
Type 2
Treatment type
Diet only
I retested because I didn’t expect it and then again when it was so different. I did already explain. The difference between 9 and 7.2 is more than 15%. (Yes I realised with the benefit of the last two that it was probably in truth somewhere in between the two). I realise tests have tolerances I just was shocked it was that high on the same monitor at the same time on the same blood with no variables applicable eg temp, blood site etc.

Once it was already higher than acceptable to me theres little point keep testing as I’d already ruled it out. I do miss bread hence trying a few different brands to see if there was one I found acceptable. Some people do seem to find some options for an odd potato, slice of bread etc. Just need to find out for myself I feel I’m one of them.
 

KK123

Well-Known Member
Messages
3,967
Type of diabetes
Type 1
Treatment type
Insulin
Have you tried using an Abbott meter and strips, which Dr Bernstein says is just about the only one that is accurate? I fear the strips are very pricey. I am listening to a podcast of Dr B's latest Teleseminar today, and he says all the other meters are grossly inaccurate. I am hoping that with the continuing development of affordable CGMs the strip sellers will have to clean up their act, as the only reason to buy their stuff will be to check the CGMs against something more accurate.

Hi Alexandra, I am using a freestyle optimum which is the one the Drs gave me, they will only prescribe the strips for that one. I am loath to have to fork out for test strips but thanks for your reply.
 

Book_woorm

Member
Messages
16
Type of diabetes
Type 1
Treatment type
Insulin
Have you got any 'Test Solutions' from Accu Chek yet? I test every new cassette that I put in my AccuChek mobile. I know this is using up one of the 50 tests in the cassette; But if you compare the test reading on the meter with the average of the high and low limits for the solution you have used it will give you an idea how far off true that particular combination of cassette and meter are at that time. As a chartered engineer the +/-15% tolerance appals me but it is the nature of the biochemistry that these test rely on it simply cannot be any better even in a lab.
 
  • Like
Reactions: HSSS

HSSS

Expert
Messages
7,465
Type of diabetes
Type 2
Treatment type
Diet only
Have you got any 'Test Solutions' from Accu Chek yet? I test every new cassette that I put in my AccuChek mobile. I know this is using up one of the 50 tests in the cassette; But if you compare the test reading on the meter with the average of the high and low limits for the solution you have used it will give you an idea how far off true that particular combination of cassette and meter are at that time. As a chartered engineer the +/-15% tolerance appals me but it is the nature of the biochemistry that these test rely on it simply cannot be any better even in a lab.

No I haven’t. Tbh I didn’t even know about them. I’ve been incredibly lucky in getting it from my DN and 2 cassettes a month. I didn’t realise I could check it. I have noticed the 4 cassettes I’ve used so far have a 2019 date stamped on the box but the validity on the meter says 11/2018. As it was still in date I didn’t worry but maybe there’s a bit of an issue. I’ll call accu check tomorrow and have a chat.
 

Mr_Pot

Well-Known Member
Messages
4,573
Type of diabetes
Type 2
Treatment type
Diet only
Have you got any 'Test Solutions' from Accu Chek yet? I test every new cassette that I put in my AccuChek mobile. I know this is using up one of the 50 tests in the cassette; But if you compare the test reading on the meter with the average of the high and low limits for the solution you have used it will give you an idea how far off true that particular combination of cassette and meter are at that time. As a chartered engineer the +/-15% tolerance appals me but it is the nature of the biochemistry that these test rely on it simply cannot be any better even in a lab.
I am also an engineer but I don't think the test solution does more than check your meter isn't "broken" rather than calibrate it. The tubs of strips for my TEE2 meter have a range like Control A: 5.9-8.9 Control B: 10.7-16.1 which is a +/- 20% range so how does that help or have I missed something?
 

Bluetit1802

Legend
Messages
25,216
Type of diabetes
Type 2 (in remission!)
Treatment type
Diet only
No I haven’t. Tbh I didn’t even know about them. I’ve been incredibly lucky in getting it from my DN and 2 cassettes a month. I didn’t realise I could check it. I have noticed the 4 cassettes I’ve used so far have a 2019 date stamped on the box but the validity on the meter says 11/2018. As it was still in date I didn’t worry but maybe there’s a bit of an issue. I’ll call accu check tomorrow and have a chat.

The validity on the meter is always a lot earlier than the date stamp on the box for the simple reason you have opened the packaging and unsealed the cassette. It will tell you somewhere in your leaflet what the time span is after unsealing the cassette. From memory I think it is 90 days, but could be wrong. There is nothing wrong with your cassettes.

Also, the control solution just checks the meter is working to within the accepted parameters and is working properly. I have only used a control solution once in 4 and a half years. They are just a ruse to get you to spend more money, in my opinion.
 

HSSS

Expert
Messages
7,465
Type of diabetes
Type 2
Treatment type
Diet only
The validity on the meter is always a lot earlier than the date stamp on the box for the simple reason you have opened the packaging and unsealed the cassette. It will tell you somewhere in your leaflet what the time span is after unsealing the cassette. From memory I think it is 90 days, but could be wrong. There is nothing wrong with your cassettes.

Also, the control solution just checks the meter is working to within the accepted parameters and is working properly. I have only used a control solution once in 4 and a half years. They are just a ruse to get you to spend more money, in my opinion.
Good to know. Thanks
 
  • Like
Reactions: Bluetit1802

Mr_Pot

Well-Known Member
Messages
4,573
Type of diabetes
Type 2
Treatment type
Diet only
Have you tried using an Abbott meter and strips, which Dr Bernstein says is just about the only one that is accurate? I fear the strips are very pricey. I am listening to a podcast of Dr B's latest Teleseminar today, and he says all the other meters are grossly inaccurate. I am hoping that with the continuing development of affordable CGMs the strip sellers will have to clean up their act, as the only reason to buy their stuff will be to check the CGMs against something more accurate.
Did he say why the Abbot meter was the only one that was accurate, If he meant the Optiom Neo it seems to meet the same accuracy standard as all other modern meters.
 

jonathanTHP

Newbie
Messages
4
Type of diabetes
Type 2
Treatment type
Tablets (oral)
Hi
If by "back to back" you mean that all 4 test was taken in a time frame of 1-2 minutes than by no means you should have got those results unless something is wrong with your meter or strips. I use the Contour Plus meter and back to back measurements always give me the same results in less than 5% range. If I understand correctly, the 15% accuracy of the meters refers to a possible calibration bias of up to this deviation but not that this is the "test to test statistical variance".
Anyway - this is my own experience.
Best,
J.