Decimal point in BG readings

Status
Not open for further replies.
H

HalfpipMarathon

Well-Known Member
Relationship to Diabetes
Type 2
Can I ask if anyone takes any notice of the number after the decimal point in BG readings ie 7.9 or 10.2? Especially the latter as 10 is the top end range for Dexcom and Libre.

I ask because if you get a reading of 10.2 rather than 10 does that mean you are not in range?

I'm not too worried just curious.
 
It is meaningless. I took a vow never to quote a single reading including a number after the decimal point after doing some reproducibility tests - sort of thing I would have done in the lab when presented with a new piece of testing kit. Round the reading to the nearest whole number and expect a 95% confidence interval of +/-2 units for any single reading is about as good as you can get and to my mind it is amazing that they get that close.🙂
 
Docb said:
It is meaningless. I took a vow never to quote a single reading including a number after the decimal point after doing some reproducibility tests - sort of thing I would have done in the lab when presented with a new piece of testing kit. Round the reading to the nearest whole number and expect a 95% confidence interval of +/-2 units for any single reading is about as good as you can get and to my mind it is amazing that they get that close.🙂
Click to expand...
Okay as I said in terms of TIR should I disregard the number after the decimal point? I go on the TIR rather than individual readings now. It's not been easy but I think I have turned a corner
 
I take a bit of notice, because I tend to think, say, 7.1 is a bit better than 7.9. But I don’t think, ooh er, 10.2, that’s out of range, any more than I’d think, Oh phew, 9,9, that’s in range. I look at the daily graph, and the weekly average graph, and if it’s pottering around in the middle, with excursions either side, I think, Oh that’s Ok, but if I was 100% in range, but the line on the graph was hovering along just below 10 the whole time, I’d think, that’s not OK.
 
Robin said:
I take a bit of notice, because I tend to think, say, 7.1 is a bit better than 7.9. But I don’t think, ooh er, 10.2, that’s out of range, any more than I’d think, Oh phew, 9,9, that’s in range. I look at the daily graph, and the weekly average graph, and if it’s pottering around in the middle, with excursions either side, I think, Oh that’s Ok, but if I was 100% in range, but the line on the graph was hovering along just below 10 the whole time, I’d think, that’s not OK.
Click to expand...
That's how you look it. I would feel although 9.9 is not brilliant it's within range ie under 10 but felt 10.2 was out of range but now think.2 is neither here nor there
 
HalfpipMarathon said:
That's how you look it. I would feel although 9.9 is not brilliant it's within range ie under 10 but felt 10.2 was out of range but now think.2 is neither here nor there
Click to expand...
Certainly for a spot reading, I wouldn’t worry if I was 10.2 as opposed to 9.9, and if the Libre was showing an upwards arrow, it may well be that the 10.2 has disappeared when I look at the graph, because the algorithm that closes up the gap between the interstitial fluid reading and an actual blood reading (about 15 minutes) might well have overreacted and assumed I was still going on rising, when I might have hit the peak and be on a plateau, or indeed a downturn. (the algorithm is often responsible for those annoying 'try again in 10 minutes' messages, when it feels it hasn't got enough data to go on.)
(I apologise to anyone of a Scientific disposition, in endowing the algorithm with human sensibilities)
 
HalfpipMarathon said:
Can I ask if anyone takes any notice of the number after the decimal point in BG readings ie 7.9 or 10.2? Especially the latter as 10 is the top end range for Dexcom and Libre.

I ask because if you get a reading of 10.2 rather than 10 does that mean you are not in range?

I'm not too worried just curious.
Click to expand...
There is so much variation between glucose meters and, even, batches of test strips that I do not think it makes sense to pay such close scrutiny. At the end of the day, we are trying to reproduce lab test results in a non-lab environment - something that is just not achievable. The name of the game in relation to blood sugar control is to focus, in my view, on the overall statistical picture rather than be overly concerned about individual readings. Certainly, for my part, I pay attention to the shape and median value in my AGP graph, along with my TIR percentage. Individual spikes and dips do not concern me too much, as long as they are not extremely out of range on either end.
 
littlevoice359 said:
There is so much variation between glucose meters and, even, batches of test strips that I do not think it makes sense to pay such close scrutiny. At the end of the day, we are trying to reproduce lab test results in a non-lab environment - something that is just not achievable. The name of the game in relation to blood sugar control is to focus, in my view, on the overall statistical picture rather than be overly concerned about individual readings. Certainly, for my part, I pay attention to the shape and median value in my AGP graph, along with my TIR percentage. Individual spikes and dips do not concern me too much, as long as they are not extremely out of range on either end.
Click to expand...
I agree. As I said in my post I am not that fussed just curious. I only look at the TIR as long as it doesn't drop below 70% or if it does not for too long I am happy.

It's a very individual thing but sometimes I read posts or article as saying it's not good for bg to be out of range or even higher than 8.5 post meals so it is understandable people can feel like they are failing if their bg levels are like this even if for a short time.

I'm coming to terms with the fact bg levels rise and fall with a lot of variance so now try to pay more attention to TIR.
 
helli said:
As an insulin sensitive Type 1, I definitely take notice of the decimal points when my levels are below about 8.
The difference between 10.0 and 10.5 is less significant for me than the difference between 4.0 and 4.5.
Click to expand...
I understand that but for me,as a Type 2 not on insulin or medication that might cause hypos just wanting to keep an eye on my bg levels I suppose it's less of an issue, however I was just curious as to whether the number after the decimal point was counted in deciding if the full reading was considered in range or out of range.
 
They are so inaccurate it's meaningless to be honest.
A reading of 10 is actually 8.50 to 11.76 95% of the time.
The other 5% it can anything at all.
 
travellor said:
They are so inaccurate it's meaningless to be honest.
A reading of 10 is actually 8.50 to 11.76 95% of the time.
The other 5% it can anything at all.
Click to expand...
If this is true why do people rely on Libre, Dexcom or any other cgm/flash gm? I know some will finger prick to check but if so unreliable why not just finger prick? Just my opinion.

I don't rely on any of them, they're just a way to ensure I am in range most of the time to not information finger pricking gives. As I self fund due to finances I don't always use them.
 
HalfpipMarathon said:
If this is true why do people rely on Libre, Dexcom or any other cgm/flash gm? I know some will finger prick to check but if so unreliable why not just finger prick? Just my opinion.

I don't rely on any of them, they're just a way to ensure I am in range most of the time to not information finger pricking gives. As I self fund due to finances I don't always use them.
Click to expand...
@travellor's comments apply to finger prick readings just as much as CGMs. The short answer is, it’s a hell of a lot more accurate than boiling up a test tube of pee with a reagent and trying to compare the colour change against a paper chart!
 
HalfpipMarathon said:
If this is true why do people rely on Libre, Dexcom or any other cgm/flash gm?
Click to expand...
Because what they give is good enough to be useful.
HalfpipMarathon said:
I know some will finger prick to check but if so unreliable why not just finger prick?
Click to expand...
Because it's really useful to see trends and patterns, and to know (roughly) what happened over night, or just in a few hours when you didn't want to test.
 
Robin said:
@travellor's comments apply to finger prick readings just as much as CGMs. The short answer is, it’s a hell of a lot more accurate than boiling up a test tube of pee with a reagent and trying to compare the colour change against a paper chart!
Click to expand...
Okay assumed he was just referring to cgms etc although I have read that finger pricking was more accurate as cgms use interstitial fluid not blood but I suppose not even that is infallible all the time.
 
HalfpipMarathon said:
Okay assumed he was just referring to cgms etc although I have read that finger pricking was more accurate as cgms use interstitial fluid not blood but I suppose not even that is infallible all the time.
Click to expand...
See here for a good article on the accuracy of finger prick meters.
jdrf.org.uk

How accurate is my glucose monitor? | JDRF

jdrf.org.uk jdrf.org.uk
If you click the link in the article 'interesting results (highlighted in blue quite far down) it takes you through to a table of results from a trial that tested all sorts of meters to determine their accuracy. Even the best ones which were 100% within the allowed 15% margin fell down when the percentage accuracy was measured at 10% and 5%.
387140A7-1604-4188-9A24-A94D81B19050.png
 
Robin said:
@travellor's comments apply to finger prick readings just as much as CGMs. The short answer is, it’s a hell of a lot more accurate than boiling up a test tube of pee with a reagent and trying to compare the colour change against a paper chart!
Click to expand...

I'm not entirely sure CGMs actually have a standard, they should have a published overall mean absolute relative difference (MARD), which is in a similar ballpark.
 
Last edited by a moderator:
Robin said:
See here for a good article on the accuracy of finger prick meters.
jdrf.org.uk

How accurate is my glucose monitor? | JDRF

jdrf.org.uk jdrf.org.uk
If you click the link in the article 'interesting results (highlighted in blue quite far down) it takes you through to a table of results from a trial that tested all sorts of meters to determine their accuracy. Even the best ones which were 100% within the allowed 15% margin fell down when the percentage accuracy was measured at 10% and 5%.
View attachment 23086
Click to expand...

I use a Contour Next - I got it for free from the manufacturer. Expensive strips!
 
Status
Not open for further replies.
Back
Top