#freestylelibre – How Libre changes glucose testing behaviour and the cost implications

#freestylelibre – How Libre changes glucose testing behaviour and the cost implications
#freestylelibre – How Libre changes glucose testing behaviour and the cost implications

Whilst responding to a comment on the Freestyle Libre in a well  known internet forum, it occurred to me that there are a couple of massive things that the little device has changed about my behaviour. There are also a couple of things that are hugely in the little devices favour that I hadn’t previously considered.

Changing behaviour as a result of using the Libre

Where to start? Well I think I’ll go with the latter first as this is what drives the former. There’s been some concern that the Libre isn’t a CGM and that it doesn’t have alarms, that you have to “flash” it and that it, well, just isn’t a CGM. It is, however, a Continuous Glucose Recording (CGR) device. This is just as valuable.

What do I mean? Well you have a record of your ups and downs, what inputs make a difference, how you respond to those inputs and you can estimate where you are going based on the graphic representation. You have all the same data that a CGM provides, as long as you scan it. It gives you the ability to learn about yourself. I’ve often said it’s a GPS and Route Planner compared to finger pricking, which is just a map. But you use it as if it was a finger prick, in that you determine when you are going to interact with it. And each interaction is easy, doesn’t hurt, isn’t messy and gives you tons of information.

On top of this, it has the benefit of not being a CGM. This means that it doesn’t alarm. Explicitly, this means that one of the key reasons that people drop the CGM systems is immediately removed. You can’t get alarm fatigue from it. It isn’t constantly telling you, very loudly, that you aren’t doing a very good job of managing your blood glucose levels. That’s what alarm fatigue mostly is. Getting fed up of having a nagging agent telling you your glucose management is crap. You already know that or you wouldn’t have a CGM in the first place. Removal of alarm fatigue means that one of the key objections to paying for CGM is also removed.

So if we put these two things together, we have most of the benefits of CGM, but it’s all done from the perspective of a finger pricking approach. You, the user, are the one in control.

Finger pricking and CGM, without the downsides of either!

This is where the behavioural change starts. We all know how much of a pain (literally and figuratively) finger prick testing can be, Most Type 1s test a reasonable amount, but that will vary day to day and there are times when you just don’t want to or it just isn’t convenient. Prior to Libre, I typically tested on average, 4-6 times per day. That’s okay, but not remarkable. When I was ill or having a bad day, it was up to 20. But that’s still only 4-6 times normally.

Moving on to the Libre, initially, I scanned all the time as it was novel and awesome. I was doing it up to 40 times per day. But I’ve been using the little device for a good 15 months now so I don’t do that any more, right? Wrong. What’s interesting is that my use of it hasn’t dropped off. At all. I still use it to scan around 40 odd times per day. Many other users report this level of “obsession” as well.

Why is this important? Using SMBG, I simply just wouldn’t have done this. SMBG is time consuming, requires a level of preparation, can be messy and generates trash that needs to be disposed of. Scanning doesn’t. It’s hard to use a blood testing meter on a crowded train, out in the rain, wandering around a festival or at a football match. Anywhere where you are getting muck of some description on your hands is likely to suffer errors.

It’s not hard to scan your sensor. It’s very easy. It’s not painful and every time you scan you don’t just see a number but an indication of what is dynamically going on. This dynamic visibility is what drives you to test again. And again. It’s a virtuous circle. And it’s not driven by being told to by an HCP. It comes out of your own curiosity and desire. It has a very different driver.

As a result, you are much more aware of the changes in your glucose levels, and when you are going up and down. For the first time, you actually do feel as though you have a level of control over what you are able to do. Control is a word oft used by diabetic consultants. It’s a very bad word with a lot of implications. But Libre changes that. It gives you control in a way that I’ve never experienced.

That’s a profound change. Would I ever have managed to maintain my blood glucose levels at the levels shown below when using SMBG alone? Maybe, maybe not. With this technology I can see it and I can do what I need to to maintain it. I can also instantaneously see the results. They have a velocity and a dynamic character. It becomes my personal goal everyday to keep inside the blue area, and in doing so, reducing my glucose level variation. It’s like a real-life computer game. It’s that different.

In the context of this experiment of n=1, the question has to be asked, “Is it only me that has seen this change?” and in conversations on facebook and twitter, I don’t think it is. With this technology available, think how you could use it in combination with a phone to encourage people to test!

Self Monitoring of Blood Glucose (SMBG)

Now referring this back to the whole SMBG discussion. The NICE Guidelines section 8.2 on page 201 recognise SMBG up to ten times per day. Reading through the tables of evidence in the appropriate documents, there is a very clear correlation between reduced Hba1C and frequency of testing. SMBG 8-10 times per day is stated as being the most cost effective approach. There is, however, a caveat that increased SMBG may reduce hypos, which there isn’t enough data to support.

I’ve also been told that data supports 10-12 tests per day as an optimal strategy, but I’ve not been able to find the reports that back this up.

The most recent study I can find looking at this is from Journal for Diabetes Investigation and contains the below regression:

While this shows a general reduction in Hba1C by increased SMBG, it doesn’t show dramatic improvements. The trend, however, is consistent with that of other reports. The limitation is that the studies only looked at up to 8 tests per day and that there isn’t a vast amount of data involved.

The shortfall in all of this is that the data doesn’t show the reduction in variance of glucose levels achieved by increasing the number of tests.

Let’s take a step back then. We’ve ascertained that increased SMBG helps to reduce Hba1C levels. We believe it also reduces the variance of blood glucose – but the evidence for this isn’t great. But we also know that SMBG is something we must do, and it isn’t all that pleasant. Many people don’t test 8-10 times per day. It’s hard work and unpleasant. Hba1Cs throughout the country would tend to agree with this.

There is unlikely to be any kind of RCT that would compare 2, 4, 6, 8 , 10 ,12, 20 or 40 tests per day as it would be an unethical trial that potentially put the participants at risk, so much of the data comes as a result of previous studies or specifically taking data in a case study like manner. We therefore don’t have visibility of what even more SMBG testing might determine. Anecdotal feedback suggests that it often results in better levels of management. We therefore simply can’t compare using the Libre up to 40 times per day with using SMBG. There just isn’t any data.

Do the costs compare?

Stepping away from the lack of comparative data, let’s then compare costs. I tweeted about saving £400 per month when doing 40 checks per day compared with SMBG. It was a hypothetical speculation based on the a back of a fag packet calculation. Looking into it in more detail, this is what I’m comparing:

Taking the NHS cheapest standard SMBG strips, the GlucoRX Original strips, they are £9.45 per 50.

The Libre sensors cost £103 per month inclusive of postage.

The assumption is that you still need to test at least once daily when using Libre. The reality is probably twice daily. Let’s assume then that we use 100 strips per month with the Libre, at a cost of £18.90, so our Libre cost works out at £121.90 per month.

If I was matching the number of tests that I do with the Libre I would be using 80% of a pot of strips a day. This would result in (based on a 31 day month) 1,240 strips a month, and before anyone says it, very sore fingers. That’s 24.8 pots of 50 per month. In the hypothetical world we are in here, that works out at £234.36 per month.

The saving of using the Libre, if we were ever doing 40 SMBG tests per day, is therefore not really £400, but nearer £112.46. Perhaps not such a big figure after all. But when you consider that significant numbers of people use much more expensive strips, cost around £16 per box, we see the savings will range from £112 at the low end to £275 at the higher end. That’s not insignificant. And we haven’t touched the costs of reduced Hba1Cs and subsequent reductions in complications and hospitalization.

I’m well aware that this is very much hypothetical, but it goes to show how changing technology can change behaviour, and what the potential knock on effects might be.

When asking for things on the NHS, as a number of petitions do, it’s wise to be aware of the impact of much more than simply “I want one. It helps me”. Including this kind of analysis in what’s being asked for makes the argument so much stronger.

“I want one free because it makes a difference to me”. There is so much more to some of these technologies than that.


  1. Last Hba1C was 39 mmol/mol (5.9% in old money). That graph was overnight though. Amazing what benefit not eating anything has on levels 😉

  2. haha..that's great.btw have u ever wondered, this Hba1c value does refer to which point in your AGP :p – some ppl say the Hba1c no.(when converted) is very similar to 2-3hrs PPBG levels.how often do you conduct Hba1c test clinically and do you think it gives a 3 month average or shorter?(like the most recent 2 weeks ave) .And from your experience, Libre predicted hba1c value, gives a good estimation about the true clinical HbA1c value? And thanks very much for highlighting abt AGP in 1 of your past post and I found this nice introductory video on this site watch@min 2 it emphasis the significance of AGP with regard to Hba1c. https://www.diabetesfrontier.us/ video 1.. whole a lot of questions 🙂

  3. Given what Hba1C is really about, I'd say that it relates to the amount of time that your ABGP spends statisitcally in relation to that of a non-diabetic. In other words, if you have a 75-percentile that regularly hits the 12s you are going to struggle to get a lower Hba1C.

    Based on observations from playing with the duration of the data, I've found the last month of data seems to be the most similar to my clinical Hba1C, but this seems to vary for different people. This could relate to whether someone is a high or low glycator as well though so isn't necessarily a clear indicator. I've been doing Hba1C six monthly, and that gives enough data so far to allow me to know which duration to use from the Libre.

  4. Yes. During the first 24-48 hours, many users have found that it can be out by quite a bit. Give it 24 hours and see where it is then. If you can get hold of some Abbott Optium test strips, you can fingerprick using the Libre for Abbott's customer services use.

    If, after 24 hours, you are still seeing a very large variation (>15%) between blood and scan, then call Abbott customer services on 0500 467 466 for the UK to identify the sensor as failing.

Leave a Reply

Your email address will not be published.