Libre users - thoughts on readings

[Matt Cycle]

I've been using the Libre for a few days now and on the whole I am finding it extremely useful. I know however several people have raised issues over the accuracy of sensor readings.

In the short time I've been using it I have noticed in general the sensor tends to reads slightly higher than a standard bg test. (There is no issue here over meter discrepancies as these are a direct comparison with an Abbott strip in the Libre reader). The other day the Libre read 13.4 but a bg test taken at the same time in the same reader showed 10.5 - almost a 3 mmol/l difference. I know there is a slight delay between interstitial fluid measurements and bg measurements but surely not to that extent. Most of the time the readings are within 1 to 1.5 mmol/l of the bg test which I would guess is acceptable but it's these very occasional rogue readings which could be an issue.

Anyone else experience anything similar to this and what would you say would be an acceptable difference between sensor reading and bg reading?

 

[Robin]

I'm finding my Libre reads lower than my Accuchek Aviva at the bottom end ( Libre said 4.2 the other day, meter said 5.2, and that was at a time when I would expect levels to be steady, so no problem with the time lag) and higher at the top, but in the middle it's spot on. I think the time lag is about ten minutes, so if you take a reading on both at the same time at a moment when your levels could be changing rapidly, I would expect a larger discrepancy than normal.
Maybe a good test would be to swipe again with the Libre after ten minutes, and see if it's closer to the previous strip test at that point.

 

[Sally71]

We've just finished our first sensor and were quite impressed! I think the biggest discrepancy we had was about 2mmol, most were closer than that and we even had one which matched exactly! On the whole we found that it mostly read lower than the finger prick test, so on a few occasions it announced that daughter was hypo when she wasn't; she knows how she feels though and if ever not sure she would do a finger prick to check. I have loved having all the extra information in the trends and graphs!
We don't have any strips for the Libre itself though, all our comparisons have been done with the Accu Chek meter which goes with the pump.

 

[Sally71]

Posted at the same time as Robin - that's a good point actually, I think the couple of times that the Libre was reading higher than the meter were at the high end. I shall take note more carefully this week and see if your analysis is true for us too!

 

[Annette]

Same here-at the low end it reads low by a point or so, at the higher end it reads high by upwards of 2 points, the higher the number, the greater the discrepancy.

 

[trophywench]

Be aware that someone complained about discrepancies and Abbott wouldn't entertain the meter readings cos it wasn't their meter. So in other words with a Roche pump - you have to do every blood test twice in order to get customer service to listen to you if you notice discrepancies and think it's a duff sensor.

This to me was a VERY black mark since I use the Insight which replaced my Combo and we have to carry more junk around with us for starters without having to carry an extra meter as well as the scanner for the Libre if you decide to invest on one. I don't suppose it calculates IoB or works out your dose either .......

Think I'll keep the money in the bank for now frankly. And that they are cutting their nose off to spite their face if they want to make it difficult for pump users, don't you?

 

[Sally71]

TW - the Libre reader is not huge so not a lot extra to carry around, we find we can fit it easily into the same bag that daughter carries her other D-clobber around in. I love having all the extra data, she loves being able to swipe instead of prick, so worth the effort in our opinion!

We are only just starting our 2nd ever sensor so no problems yet, if we do ever get a duff sensor and they get a bit arsy about which meter we are using, well I'll try the trick of blagging a box of their strips out of them and then use the Libre as the "appropriate" test meter. If I have to buy them then one box is hardly going to break the bank. If you're doing finger pricks anyway (which we still do at mealtimes to get the BG into the pump meter so that it can work out the dose) then it's easy enough to dip two strips into the same blood drop.

I believe there is a bolus calculator in the Libre but it sounds a bit simplistic and too complicated to change the settings; all pumps do that for you though I think. We are up for new pump in the summer, if we choose anything but the Insight then I think I'll just get the prescription changed over from Accu Chek to Abbott strips anyway and use the Libre as the blood meter too if we want to check the sensor, then it's only one meter for daughter to carry around.

Being as we want CGM data but have to self-fund, and this is the cheaper option, it's not so bad I reckon!

 

[DeusXM]

I believe there is a bolus calculator in the Libre

There is, but it doesn't actually work if you just use the scanners. You can only use the bolus calculator if you use the strips in the meter too.

That should tell you something about the Libre's accuracy - evidently, Abbot don't really have enough faith in the scanner's accuracy to provide bolus guidance.

Having said that, I personally have found the Libre to be pretty much bang on, particularly in the all-important 4-7 range. It is by no means a perfect device but I find that it is at least consistent, which means you can certainly make use of the trend graphs for predictions. That for me is the real benefit from the Libre - I'm actually less bothered about how accurate it is, provided it gives me that kind of trend data.

 

[jusme]

I find the Libre useful for checking BGs during the night without having to get out of bed to go to a room where I can turn the light on and wash my hands to do a BG test and it is also useful during the day seeing what is going on whether I am busy at home or out shopping. I tend to only use Bayer Contour before meals and before bed or if the Libre is showing too high or too low a reading. After doing finger pricks since BG meters were available it's nice to cut down and give my fingers a rest.

jusme

 

[everydayupsanddowns]

The other day the Libre read 13.4 but a bg test taken at the same time in the same reader showed 10.5 - almost a 3 mmol/l difference. I know there is a slight delay between interstitial fluid measurements and bg measurements but surely not to that extent. Most of the time the readings are within 1 to 1.5 mmol/l of the bg test which I would guess is acceptable but it's these very occasional rogue readings which could be an issue.

I think it is useful to bear a few things in mind when considering Libre and accuracy.

The first is the assumption that the BG meter is 100% right all the time. Under the new (tighter) ISO regulations meters have to give results within these accuracy guidelines 95% of the time - but can read anything at all for the other 5%:

You can see that up around the 10-13 level the results really start to spread out and overlap and unless you want to take every single BG meter reading 2-3 times and find an average I think you just have to accept that all these numbers are very useful and can be used to inform your decisions, but differences between devices and technologies are pretty much inevitable.

The other thing, as you have suggested is that interstitial fluid and BG meters are reading completely different things. The Libre takes raw data and then interprets it through an algorithm to try to close the 10-15 minute lag to more like 5 minutes. This means that it takes direction of travel into account and can struggle where BG curves are turning or fluctuating. Abbott are quite upfront about this in the manual.

By and large I find the Libre very reliable within the 4-8 range (most often within 0.5-1.0 mmol/L) Once up above 10 I worry less about a bigger difference (as it represents a smaller proportion) and I know I need to do something about it anyway.

I can understand your concern that you might have overdone a correction if you dosed based on 13 rather than 10 (potentially a whole unit you didn't need) but my experience is that corrections, as well as BG readings, are very much affected by 'direction of travel' and at least with a Libre I can watch what's happening in the meantime.

If I get a sensor that consistently reads >2-3mmol/L out at all levels I would be tempted to contact Abbott and ask for a replacement, but if it's 2mmol/L out as a one-off when I'm 10 and rising I'm more relaxed about it personally.

If you are interested, I did my own compare-and-contrast between Libre and BGs when I first tried the system just before it was launched: http://www.everydayupsanddowns.co.uk/2014/10/abbott-freestyle-libre-results-vs-bg.html

 

[HOBIE]

A very good gadget ! When you see the graph of when you are asleep etc 😎

 

[jusme]

Now, now Hobie I am sure you know what I meant and by the way I have edited my reply to make it clearer for you.🙄

jusme 🙂

 

[Squirrel768]

Hi, Recently started using Libre. As a newbie to it, and maybe too much of a geek by nature LOL, I've been covering all bases, using the mobile app mainly, along with the reader and finger prick tests particularly at breakfast, to give me a baseline comparison with the 'old' regime (Contour Next One ).

So far, results in general are roughly as follows, with a small number of variations, and regardless of which direction the trend is going.

Reading on the app is, usually, lower than on the reader, mostly between 0.5 and 1.0 mmol. Occasionally the same, and rarely lower on the reader.

Reading from finger prick is higher than both app and reader, with rare occasions when it is in between, or same as the app.

So, the app can show a hypo, when the reader is 'safe', and blood test is higher still.

So far, on my 3rd sensor (first one was swapped after a week and a half following a CT scan). Different sensors behave slightly differently, but all seem consistent in terms of the trends etc.

Not sure how this will go long-term, but I'm certainly very pleased with seeing the trend graphs (and picking up on regular peaks and dips), so looking forward to improving my care in the future. Been pretty well controlled for 32 years on insulin, so hope it will be as good for the future decades 🙂

 

[Amity Island]

I've been using the Libre for a few days now and on the whole I am finding it extremely useful. I know however several people have raised issues over the accuracy of sensor readings.

In the short time I've been using it I have noticed in general the sensor tends to reads slightly higher than a standard bg test. (There is no issue here over meter discrepancies as these are a direct comparison with an Abbott strip in the Libre reader). The other day the Libre read 13.4 but a bg test taken at the same time in the same reader showed 10.5 - almost a 3 mmol/l difference. I know there is a slight delay between interstitial fluid measurements and bg measurements but surely not to that extent. Most of the time the readings are within 1 to 1.5 mmol/l of the bg test which I would guess is acceptable but it's these very occasional rogue readings which could be an issue.

Anyone else experience anything similar to this and what would you say would be an acceptable difference between sensor reading and bg reading?

Hi Matt,

I'm still in awe of the libre, even after using it for nearly 2 years. It's made life so much more enjoyable, fullfilling, safer and easier in every way one could imagine. How often in life do you try something, use something, eat something that goes way beyond your expectations? It's very rare. The libre is one of those things where you're blown away by it.

I've found the lows are lower than the accuchek meter and the highs are usually higher. I've also found sometimes the libre reads hypo (below 4) but blood test is still in range. These quirks with the libre don't concern me, as I am just so grateful to have it.

 

[Bruce Stephens]

Reading on the app is, usually, lower than on the reader, mostly between 0.5 and 1.0 mmol. Occasionally the same, and rarely lower on the reader.

Is that the official app? That seems odd: I find they're the same (maybe .1 different if I'm a bit slow between the two). They're both reading stuff from the sensor and (as far as I'm aware) using the same algorithm to adjust (I'm not sure what that involves but I'm guessing not that much).

 

[Squirrel768]

Is that the official app? That seems odd: I find they're the same (maybe .1 different if I'm a bit slow between the two). They're both reading stuff from the sensor and (as far as I'm aware) using the same algorithm to adjust (I'm not sure what that involves but I'm guessing not that much).

Hi Bruce - yes, the official download from Playstore on Android/Samsung Galaxy. Now on my third sensor, and so far there seems to be little difference in the outcomes ... I've been puzzled by the difference between app and reader with no more than a few seconds between scans.

 

[grovesy]

Is that the official app? That seems odd: I find they're the same (maybe .1 different if I'm a bit slow between the two). They're both reading stuff from the sensor and (as far as I'm aware) using the same algorithm to adjust (I'm not sure what that involves but I'm guessing not that much).

I find app and reader the same.

 

[trophywench]

Ooopsie- just scanned - 9.0 with a horizontal arrow. Perchance I should have bolused for the slice of Tesco's Genoa cake getting on for an hour ago ...... Ho hum, this is what I do more than on the odd occasion.

9 minutes later (cos I needed the loo) BG meter said 11.4 so I've had a correction and then just for the hell of it scanned again - 10.4 and upward arrow. Does not exactly compute in terms of arrows anyway. That's a bit odd.

 

[helli]

@trophywench I find the arrows pretty useless because they only seem to indicate the change between the last two values. Often Libre seems to have one or two values which do not follow the general trend meaning the arrows point in the overall wrong direction as you have seen.
I always look at the graph to see the trend

 

[everydayupsanddowns]

Hi Bruce - yes, the official download from Playstore on Android/Samsung Galaxy. Now on my third sensor, and so far there seems to be little difference in the outcomes ... I've been puzzled by the difference between app and reader with no more than a few seconds between scans.

I find app and reader the same.

I agree with @grovesy - App and reader should be the same as far as I am aware.

I know you have said you've started using recently, but is there a chance your reader is still using the older algorithm? That’s the only reason I can think why they may be different unless you are timing scans just either side of the sensor datapoints?