Libre v Accu-chek Expert - personal experiment

[Anthony_H]

Hi all,

I’ve been injecting insulin now for just around 5 weeks after being diagnosed 24th August.

On Tuesday I was ‘fitted’ with a Freestyle Libre and I’m trialling it for two weeks.
I thought I’d be pretty impressed with it after reading numerous forum discussions relating to it but so far I haven’t found it as useful as I thought it would be and certainly haven’t gained much knowledge about how my blood works throughout a normal days routine.

Some forum posts have made particular reference to the differences in BG test readings between the Libre v a finger pricking meter. Some unimpressed members discuss the diversity of BG levels between meters whilst others with seemingly more experience of blood glucose levels in general suggest the different reading fluctuations are not just down to the meters accuracy. Our body chemistry and varying temperature, and the area the blood sample is taken from are just some of the factors that can contribute to the variations in our BG results.

So I decided to run a number of tests just to see if I could see a pattern, all done in as quick a time as I could.

Note that just 30 minutes beforehand I had injected 6 units of rapid acting insulin to cover my breakfast which I injected post meal. The breakfast totalled 60g of carbs - 4 weetabix + 1/3 pint semi skimmed milk.

FYI, the Libre is fitted to the back / side of my left upper arm. My finger prick meter is the Accu-chek Aviva Expert.
I performed a controlled test on both meters prior to directly comparing the results which commenced from 8.00am onwards.

He are the results for the whole of today beginning after I woke up, got dressed and went downstairs for coffee.



5.26am Libre = 8.7 375g mixed (other) medication taken in capsule/tablet form

6.21am Libre = 9.4.

7.28am. Accu-chek = 8.7 Bolus x 6 units
for 60g carbs
7.35am. Libre = 10.0 Start eating breakfast

7.55am. Libre = 10.1 Accu-chek = 12.5. Pin prick from thumb (cold from outside)

7.58am. Libre = 10.1 Accu-chek = 10.7. Pin prick from side of warm big toe

7.59am. Libre = 10.2 Accu-chek = 9.2. Pin prick from bottom edge of palm

8.01am. Libre = 9.9 Accu-chek = 13.0. Pin prick from thumb (warmed on AGA top 60 sec)

8.06am. Libre = 11.0 Accu-chek = 13.3. Bolus x 1 unit correction Pin prick from thumb

8.22am. Libre = 12.7

9.00am. Libre = 19.0. Accu-chek = 16.1. Bolus x 3 units for
17g carbs + correction
10.15am. Libre = 13.3. Accu-chek = 11.0


Now, I appreciate my overall average is relatively high but perhaps it shows the reading variations more clearly. The biggest difference is at 9am with a variation of 15% between the two meters, so both within the tolerances set by the medical conformance bods if I’m not mistaken.

From Saturday morning to Tuesdays hospital appointment earlier this week I had readings from 5.1 lowest to 12.5 highest with a target range set of between 6-10. I was within target for more than 50% of the time and consequently the staff decided not to change my ratios or basal dosages for now. My ratios are 1:10 breakfast, and /1:12 lunch & dinner. Snacks during time blocks are at the same ratios. My basal insulin is 28 units taken at 8pm. My HBA1c pre treatment was 147 and BG averaging mid to high 20’s for first 2 weeks before gradually dropping to an average of just below 10.

Hope you find this interesting.

Regards
Anthony

 

[Deleted Account]

I trialled the Libre last year and, like you, I was disappointed in the variation from finger pricks.
Since then, I have had some time to consider Continuous Glucose Monitoring including Flash monitoring like the Libre. As a result, I have decided the value in CGM and Flash is not the absolute readings but the historic data which gives us trends. I would say using something like the Libre as a finger prick replacement to give a value at a specific time is analogous to using a Smart phone just to make phone calls. This is not getting the value out of the Libre.

Incidentally, I blog for Diabetes UK. Recently, they asked me to write a post about the Libre following the recent announcement that it will be available on the NHS from next month. I explained my thoughts with this post: https://blogs.diabetes.org.uk/?p=9247

 

[Anthony_H]

I was also showing how, even with the same meters there is the chance of a relatively significant difference in our BG readings from differing parts of our body, and the temperature of those parts at the time of testing. See, for example, the test taken from my palm compared with my thumb, 9.2 v 13.0 both using the Accu-chek meter 2 minutes apart.

 

[Anthony_H]

I trialled the Libre last year and, like you, I was disappointed in the variation from finger pricks.
Since then, I have had some time to consider Continuous Glucose Monitoring including Flash monitoring like the Libre. As a result, I have decided the value in CGM and Flash is not the absolute readings but the historic data which gives us trends. I would say using something like the Libre as a finger prick replacement to give a value at a specific time is analogous to using a Smart phone just to make phone calls. This is not getting the value out of the Libre.

Incidentally, I blog for Diabetes UK. Recently, they asked me to write a post about the Libre following the recent announcement that it will be available on the NHS from next month. I explained my thoughts with this post: https://blogs.diabetes.org.uk/?p=9247

Hi Helen, ...that’s a good read, and I see from that where the benefits can be found, particularly in relation to the differing food stuffs and how ‘spiking’ could be minimised by analysing information the Libre, and other similar devices, can provide if you have the dedication to work with it. I have a food diary so I’ll sit down ove4 the weekend and see if anything jumps out at me, though perhaps it may take a little while longer before I have enough data to compare?

 

[tim2000s]

Now, I appreciate my overall average is relatively high but perhaps it shows the reading variations more clearly. The biggest difference is at 9am with a variation of 15% between the two meters, so both within the tolerances set by the medical conformance bods if I’m not mistaken.

Hi Anthony, I'd argue that your testing isn't really all that relevant, as your libre remains in a static location throughout the tests, while your blood tests don't.

If you want to make a comparison, in the below, your thumb is the only site that you could consider to be similar to the Libre and it is consistently no more than 0.43 mmol/l from the average, or in other words, has a maximum variance of 3.3%, meaning that temperature makes little difference.

Of the other sites, the toe is not approved for testing using Aviva, and the palm is when using a different fingerpricker cap, as depth affects the blood extracted.

It's well known that multiple pricks from different fingers will result in a reasonable variance in results as blood is not homogenous.

7.55am. Libre = 10.1 Accu-chek = 12.5. Pin prick from thumb (cold from outside)

7.58am. Libre = 10.1 Accu-chek = 10.7. Pin prick from side of warm big toe

7.59am. Libre = 10.2 Accu-chek = 9.2. Pin prick from bottom edge of palm

8.01am. Libre = 9.9 Accu-chek = 13.0. Pin prick from thumb (warmed on AGA top 60 sec)

8.06am. Libre = 11.0 Accu-chek = 13.3. Bolus x 1 unit correction Pin prick from thumb

With regard to the Libre, you don't need to look back over the data to see whether food causes spikes. That's the huge benefit of it. A quick scan after lunch will reveal issues in that respect.

 

[Snapsy]

I'm a huge fan of the Libre for the trend line, and the arrows are also incredibly useful, particularly when exercising. I can stave off impending hypos and hypers, as I can see them coming. It's an excellent additional tool in my diabetes armoury, and used in tandem with my pump I feel really confident in my diabetes management.

Loved your blog post, @helensaramay - I particularly liked the Nokia 3110 analogy! In the picture the sensor is being worn on the side of the arm, rather than the back - I only ever use the back of my arm for fear of bumping the sensor into things - and obviously they're expensive to replace if they get knocked off. Did you ever have this problem?

 

[Deleted Account]

In the picture the sensor is being worn on the side of the arm, rather than the back

Actually, that's not my arm ... although I love the tattoo. Diabetes UK provided the photos as I am too shy to have much more of me than one picture of me hiding behind my sunnies.
But I take your point about location of the sensor.

 

[Scott-C]

Forgive me if I keep on mentioning this (have mentioned it in a few other posts), but on the accuracy issue, I'm kinda like a 5 yr old kid with a new Tonka toy at the moment, so just wanted to mention that for the last ten days, I've had a Blucon transmitter from Ambrosia Systems sitting on top of my libre sensor, bluetoothing results every five minutes to xDrip+ on my phone.

The significance of that is that xDrip+ allows you to calibrate against bg readings, and gives hypo/alerts.

I've not yet sat down and jotted down readings in a systematic way, but my provisional view is that if calibrated a couple of times a day, results are usually only 0.2 or 0.3 out, although it is more marked during unstable periods, but that's probably more to do with biology than anything.

Hypo alerts are the main thing missing from libre. If you're not inclined for whatever reason to use dexcom, for example, you don't necessarily want alarms all the time, just now and then, or just overnight, I reckon blucon/xdrip+ fills a gap by providing hypo alerts at a modest cost, just over £100 for the blucon transmitter.

The makers say it'll last for a year, but it's just a bit of electronics so I suspect it'll last longer. It's not waterproof so needs to be taken off for showers which is a bit of a faff, but easy enough by just using an armband or plaster.

 

[therower]

Hi @Scott-C . I'm a complete techno phob, and have to be honest when I say your post above leaves me totally baffled.
On saying that I appreciate all the time and effort you've been putting in to setting up the system and how it will undoubtedly help others. Keep posting how it's going, I don't understand the lingo but it's obvious people need to know what you're doing and the benefits it offers.

 

[rider]

I'm a huge fan of the Libre for the trend line, and the arrows are also incredibly useful, particularly when exercising. I can stave off impending hypos and hypers, as I can see them coming. It's an excellent additional tool in my diabetes armoury, and used in tandem with my pump I feel really confident in my diabetes management.

Loved your blog post, @helensaramay - I particularly liked the Nokia 3110 analogy! In the picture the sensor is being worn on the side of the arm, rather than the back - I only ever use the back of my arm for fear of bumping the sensor into things - and obviously they're expensive to replace if they get knocked off. Did you ever have this problem?

Hi Snapsy, knocking it off is a smaller worry even when on the side. Have had it for about a year and the only trouble I've had is the very first one coming off due to oily skin. Since then, I have always put a waterproof skin on top of it for additional support and if I am travelling, then a nice exercise tape over it. Makes it look funky.

 

[rider]

Forgive me if I keep on mentioning this (have mentioned it in a few other posts), but on the accuracy issue, I'm kinda like a 5 yr old kid with a new Tonka toy at the moment, so just wanted to mention that for the last ten days, I've had a Blucon transmitter from Ambrosia Systems sitting on top of my libre sensor, bluetoothing results every five minutes to xDrip+ on my phone.

The significance of that is that xDrip+ allows you to calibrate against bg readings, and gives hypo/alerts.

I've not yet sat down and jotted down readings in a systematic way, but my provisional view is that if calibrated a couple of times a day, results are usually only 0.2 or 0.3 out, although it is more marked during unstable periods, but that's probably more to do with biology than anything.

Hypo alerts are the main thing missing from libre. If you're not inclined for whatever reason to use dexcom, for example, you don't necessarily want alarms all the time, just now and then, or just overnight, I reckon blucon/xdrip+ fills a gap by providing hypo alerts at a modest cost, just over £100 for the blucon transmitter.

The makers say it'll last for a year, but it's just a bit of electronics so I suspect it'll last longer. It's not waterproof so needs to be taken off for showers which is a bit of a faff, but easy enough by just using an armband or plaster.

Scott-c, that sounds amazing. Awesome work there.

I am t2 and use libre just to keep things in check. The struggle is - libre sensor has a standalone app and no way I have been able to get the readings into Samsung health. Let me see if your solution gives me any direction.

 

[Scott-C]

@Anthony_H , I think those results show your libre is actually working quite well!

You started eating about half seven, A shows 8.7, call it 9, L says 10. Those are in the same general ballpark - they are basically the same number.

Switch to 8am, L about 10, A is 13. You might regard this as a wide divergence, but it's not surprising as you've just had 60g carbs 30 mins earlier, so some of that will have already started digesting, it'll be in your bloodstream long before it's in your interstitial fluid, so it's not a surprise that libre can't measure glucose which isn't actually in the interstitial fluid yet.

The 8:06 and 8:22 readings show it is catching up.

Now, the 9:00 reading is where it gets interesting. L 19, A 16. This is an hour and a half after your bolus, so that'll definitely be at work in peak time, and you're recently dx'd as well so there's maybe a bit of honeymoon in play. What L being substantially higher than A says to me is that the insulin is now hard at work bringing your levels down in blood, and again because of the lag, it takes a bit of time for that to be reflected in interstitial fluid. I've had numerous situations where I've been swithering about taking a correction dose which can lead to awkward stacking situations, but have chosen not to because the divergence shows that the last shot is actually doing its job and I just need to be patient. These differences can actually be useful in that sort of situation. When I'm having doubts about a stubborn high and might otherwise be tempted to go on a "rage bolusing" spree, that's one of the times when a cross check between meter and libre can be useful.

I wonder sometimes whether it would be useful for newly dx'd to go back to the good old days and try colour changing strips for a while. I was dx'd in 1988 and the only thing then was putting a drop of blood on a strip, leave for a minute, wipe off, leave for another minute, then compare the colours to a chart on the side of the pack:

No such things as decimal points, treatment advice was, "try to keep between 4 and 7, have some sugar if you're below 4."

Even though I love what we've got now compared to that, I still try to keep in mind that all I really need to know is am I generally between 4 and 7, or below 4 or above 8, 9, 10. That way, I don't need to fret too much about differences - as long as I know the general park I'm in, it's fine.

For example, in the numbers you gave above, half seven, A 8.7, L 10, sure, those are different numbers, but they're both not in 4 to 7, so irrespective of which one is correct, I'd be thinking of pre-bolusing with a correction added on instead of doing the correction an hour and a half later.

Best of luck, mate, we're having to deal with a constantly moving target here! As you learn more about T1 and libre, you'll start finding uses for it which might not be apparent at the moment.

 

[barrym]

Scott-c, that sounds amazing. Awesome work there.

I am t2 and use libre just to keep things in check. The struggle is - libre sensor has a standalone app and no way I have been able to get the readings into Samsung health. Let me see if your solution gives me any direction.

If you login to the LibreView cloud store which is updated automatically you can export the BG data to a CSV file which your other system may well be able to import.

 

[Snapsy]

Hi Snapsy, knocking it off is a smaller worry even when on the side. Have had it for about a year and the only trouble I've had is the very first one coming off due to oily skin. Since then, I have always put a waterproof skin on top of it for additional support and if I am travelling, then a nice exercise tape over it. Makes it look funky.

I've never lost one yet, thank goodness - happily where I position it it seems to be safe from door frames etc, which is why I wondered about what can happen when it's applied elsewhere. I do make sure I am extra careful when getting in and out of swimwear and running vests!

 

[Scott-C]

Hi @Scott-C . I'm a complete techno phob, and have to be honest when I say your post above leaves me totally baffled.
On saying that I appreciate all the time and effort you've been putting in to setting up the system and how it will undoubtedly help others. Keep posting how it's going, I don't understand the lingo but it's obvious people need to know what you're doing and the benefits it offers.

Believe me, mate, modern phones scare me - I'd been using the same phone for 8 years and only got a new one because the back of it kept falling off. Took me a while with the new one to figure out I had to slide the answer button across to actually answer a call!

It's really not my work - am just pointing out what others have done.

It's a great story - you've got Abbot, an 88 billion dollars market cap company making libre, but no alerts, Ambrosia Systems, a tiny start-up whose registered office is a suburban mom and pop house outside San Francisco, making blucon to send results to a phone, so no scanning...but still no alerts, then the final piece of the jigsaw is jamorham tweaking his experimental xDrip+ to get readings from blucon and then stitch the three together and hey presto...alerts, calibration.

Don't know if the poster osodebailar in the link below initiated this by her change request or if jamorham was working on it already, but, at the end of the day, through an amalgam of a multi-billion dollar company, a small start up, and a genius open-source programmer, we've got libre alerts for 100 quid!

I love the way that the change request is made 3 months ago, and a couple of months later, it's like, aye, ok, done it. If that was happening in a formal company setting it would have taken yeeeeeeeaaaars...

Don'tcha just love the free world!

https://github.com/NightscoutFoundation/xDrip/issues/144

 

[Anthony_H]

@Anthony_H , I think those results show your libre is actually working quite well!



Best of luck, mate, we're having to deal with a constantly moving target here! As you learn more about T1 and libre, you'll start finding uses for it which might not be apparent at the moment.

Hi Scott-C
Thank you very much for your awesome reply. To know there are members out there who have the knowledge, and patience to respond in such detail is so useful to a newbie such as myself.....and you’ve explained things in such an easy and uncomplicated way that just makes (common) sense.
Thanks again

Anthony