In all the noise around EASD, you may not have noticed that Microtech have launched a new CGM.
This is their follow up to Aidex, and in its native market, the unit looks a lot like the Libre3. In fact, in the picture below, where it’s sitting next to a Libre2 Plus, it’s pretty much indistinguishable from the Abbott product.
According to various sources, for European launch, it will have a different shape, to avoid the obvious issues there are with Abbott trademarking a circular sensor.
It’s not been possible to find any documentation that might be considered an accuracy study, so I have no idea about any claims related to Linx performance or MARD.
I’ve been running this sensor alongside the Libre2 Plus and a regular Dexcom G6. The results have been rather interesting.
Timezone management
This model was bought internationally and imported, and I discovered, doesn’t really like timezone changes. Early in wearing it, I had to take a day trip to Paris from London.
When the clocks went forward, this happened:
As you can see, the trace goes somewhat crazy. And backwards in time.
Coming back, we got the following:
This time, the traces overlay one another as the clock changes.
My initial impression was that someone had cocked up the app design, but I’ve subsequently heard that it is designed to use UTC times and display all user data.
Clearly that doesn’t work properly, as the first image shows duplication of data in the past. It’s something I feel they need to fix. You might also notice a propensity of lows in some of these images, and that brings me on to the next point.
Inaccurate readings on the low side
The next three images show the time in range for the Dexcom, Libre and Linx over the ten days of the Dexcom.
These three images show rather different things.
The Linx suggests that I spent 25% of my time below 3.9mmol/l/70mg/dl.
It is wholly incorrect.
The Libre2+ didn’t do much better.
The Dexcom is pretty much on point.
The issue that this presents is that in a clinical setting, it would drive considerable anxiety and potentially result in changes to treatment that were incorrect and unnecessary.
It can be seen in the images below which show the three sensors side by side. The Linx recorded lows, the Libre recorded just about okay, and the Dexcom flatter and higher.
Again, at the current time in all three, the glucose was aligned, but when it was at lower levels, the Linx went very low.
This was not a good outcome for the Linx. But it wasn’t just on the low side. After calibration, it then overshot on the high side. As the next two pictures show.
The challenge here was that when I calibrated with a finger stick, it seemed to shift whole curve up and down the y-axis without taking into account anything else. This resulted in roughly okay values in the normal range but then produced either overly high or overly low values outside that.
I’m told that the calibration mechanism they use is wholly mathematical and didn’t take any kind of sensor behaviour into account, but whether that’s the correct way to do it remains to be seen.
This obviously creates challenges in terms of treatment decisions, and I wouldn’t want to use it for that, even though it is indicated as such.
What about that similarity to Libre3?
It looks almost identical to a Libre3. Sensor application is very similar to a Libre3.
And weirdly, the TIR results were remarkably similar to the Libre2 Plus, which uses the same sensor tech as Libre 3.
I don’t know whether that’s a coincidence and the fact that having them next to each other was the cause of that similarity, or whether there’s something else going on.
It’s certainly a massive coincidence…
This isn’t what Europeans are getting though?
No, it isn’t. Due to trademark infringement risk, Microtech will be producing something that’s a different shape to the Chinese and International product for Europe.
I’m told there are also differences between the Chinese product and the currently shipped international one as well, but have no idea what those are.
But would I buy the European product and trust it for dosing? Not at this stage. I’d want to see how it performs alongside other products first.
“in a clinical setting, it would drive considerable anxiety and potentially result in changes to treatment that were incorrect”
Which kinda raised the question for me of whether it’s the clinical setting they’re really aiming at. There’s been such a flurry of CGM tech designed for the consumer fitness market recently that I wonder to what extent the development of real medical devices is able to remain completely free of that influence. Most of it, where not actually fraudulent, doesn’t have anything like the accuracy required for FDA approval as a genuine therapeutic device. Like results you can dose off of or integrate into an AID system. But that’s not what those manufacturers are aimed at. They dream of an AI-driven killer-app that would function as a general-purpose health-assistant built into their ecosystem. Key thing is that your casual fitness buff doesn’t mostly want something that sticks through their skin, so all of the ones I’ve seen are non-intrusive. Which would be nice if you could get the accuracy needed for making dosing decisions and AID basal control, but that goal remains elusive.
Anyway, I’d \be interested in your thoughts on how these two markets relate and whether we’re starting to see more crossover particularly among new manufacturers.
Given that Microtech also make the Equil patch pump system, I’m fairly sure they’re a domestic Chinese company that deals in type 1 diabetes and is looking to expand their offerings globally. I wouldn’t be surprised if they have an AID algorithm hanging around somewhere.
There are some, like Sibionics, that seems to do lifestyle more than diabetes, but most of the lifestyle market stuff (Stelo and Lingo) has come from the big players.
The area it will get serious in is if anyone manages to successfully manage non-invasive effectively.
Thanks for the informative article and reply!
Very comprehensive, thanks. I’m a Libre 2 user who is frustrated with the inaccuracy and latency associated with the device when managing my levels. I find myself constantly doing blood tests to get the correct values and base my diet, exercise or insulin amounts on a combination of readings!