Rise of the Machines

So it’s the RotM (Rise of the Machines) day coming up this weekend and I’ve been looking forward to it. In typical fashion, there has been a little snag though!  

 The roads up here have been closed. To get to the event on time means a 68-mile diversion to get to the train station (and the same to get home). So that means a very silly hour start to get there with a lift. If I went by bike I could avoid most of the diversion, but that still leaves 22-miles. Thus meaning starting at about 4am, maybe sooner to be safe due to darkness/weather. Taking Lancet (the bike) would be a good thing as she is my APS system and quite frankly the sexist APS builds out there (in a Marty McFly voice “You built a APS out of a bicycle?!” Yep, I did as it has the additional sensors such as power output, heart rate, calorie burn, environment etc.)! So, it would solve both things but would require a silly hour ride in -3C/-4C temperatures through a Met Office weather warning just to get to the train station in pitch black darkness. Yes, I have lights, good lights but night riding isn’t going to be fun in those kinds of conditions. So after a fair bit of considering, sod that for a plan, I ain’t going! 

 This could have been got around by going down a day earlier and coming back on Sunday. Snag on that would mean this would soon become a £250-300 trip and I would be losing a day off work as well. So yes, a possible but a bit of a hmm is it really worth it?! 

 Anyhow, I had a few questions and points of discussion for the event though, which I’m going to throw out there so anyone can pick up and try and answer for me. Yes, in my inevitable style some sound like I’m treading on toes again but not intending to do so, would just like to know the answers to these is all and throw the ideas out there since I can’t make it… 

 

Real-world v Certified Usage 

 This is an interesting one as it’s one questioning how the NHS thinks things through. I can remember the first time I got to play with a CGM back in 2010, the *shudder* Medtronic harpoon. Evil thing that it was! It was certified for 6 days but my DSN said straight off, if it fails on day 5 give Medtronic a call and get a new one and if it gets to the end of the 6 days, restart it and see how long you can get out of it. I liked it for it being ohh looksee what my BG is doing, but soon saw the issues it had with dehydration and other problems. It was an interesting starting point though. But even in those early days a good DSN knew it’s limitations and how to flex them past their manufacturers certified boundaries also. 

 It has been shown recently that part of the issue with getting the NHS to provide CGM is money. Yes I know it was only September when we were told money wasn’t an issue as the Diabetes budget is so vast it doesn’t matter and being able to spend on the Libre wasn’t a concern. Well now it is and thus the reason why we can’t fund CGM. What was it I said at that time? Something about how the Libre would actually hinder our care by stopping us getting devices which can provide a far greater benefit (CGM as needed for use with APS systems) by stopping the budget for CGM? Anyhow, that is the past even if my Nostradamus thoughts have all come to pass or are skipping along quickly towards them. 

Part of the issue with certified usage is a Libre sensor is certified for 14 days. A Dexcom sensor is certified for 7 days. Therefore, in certified usage the Libre is a more economical device. Snag is being that the Libre isn’t restartable, at all while the Dexcom G4/G5 sensors are, multiple times. It’s not certified for this but every diabetic using one, every online forum and most health care professionals all know it. How many times? This comes from how it is used but depending on how clean the site was before the sensor was added and if some good fixing tape is used when it is applied most users get anything from 3-5 weeks with some getting up to 8 weeks from a single sensor! I know myself I average 4-5 weeks (only had once last a poor 3 weeks) on a sensor and when it comes to comparing pricing of the systems that actually alters matters greatly! 

  • The Libre using certified usage costs £1060 for the first year.
    Outside of certified usage it still costs £1060 for the year (costs include original starter kit, £990 per year in following year).  
  • The Dexcom G4/G5 for certified usage is £2900 for the first year.
    Although if you go outside of certification and restart sensors and change transmitter batteries yourself, that cost reduces to £820 for that year (price includes original starter kit £620 for the following year without this). Even if you don’t change transmitter batteries yourself and purchase new ones this equates to £925 per year. 
  • Medtronic Enlight certified usage is £3440 for the first year (excluding cost of a pump which is also necessary as it needs a Medtronic pump to display readings).
    Outside of certification this does drop to £2200 (no starter kits etc to not need calculating for further years). 

Outside of certified usage, in real world conditions a
Dexcom 
G5 is considerably cheaper than a Libre! 

The Dexcom sensor technology is considerably the cheaper option to run. Yes, I know manufacturers and the NHS can’t use anything not certified for their calculations and as such need to ignore how users actually use them. But the reality is, the Dexcom G4/G5 is the cheapest solution available, and the soon to launch G6 is expected to work out even cheaper due to sensors lasting even longer (without costs of sensors at this time pricing can’t be done).   

 

Software 

There is a lot of discussion these days about AndroidAPS, and OpenAPS which is where it all comes from. There has also been talk of a possibility of the NHS certifying it in some way. While yes, a wonderful idea it worries me in certain regards. I keep hearing the argument that the release version of the software should be the one certified by the NHS. This is an opensource piece of software, releases can come every few days or weeks. Certification takes testing, extensively, and could never keep up with that. That alone would require some considerable resources by the NHS. Simply put, never going to happen!  

The “we do testing in the dev release, the release one is perfect” argument is total utter crap! Yes, I totally understand, you spend weeks, months, years working on your code and feel it’s utterly perfect. You and your friends have spent weeks testing it and it works perfectly as you send it to become the release version. It then fails, often! It happens to all big software companies; how many times has Apple screwed up a patch? or Microsoft? Or even huge opensource projects like WooCommerce? Only a few weeks ago it’s latest release version crashed and burned (a lot!) of websites which needed rolling back until it got fixed. This was after a lot of testing also. I know myself, I’ve just earlier this month released a major project for work which I’ve spent the last 18 months on. We’ve been testing it in house and with third parties for the last few months, it was great…I spent the first few nights up to silly hours sorting all kinds of issues which came up once the general public got their hands on it. 

We all take great pride in our work, we all think it is wonderful, but we all make mistakes and when that work we are doing is put into the hands of people whose lives depend on it, we need to get it right! My proposal recently for this was a simply one, basically something akin to a LTS (Long Term Support) build for AndroidAPS and OpenAPS. This is used on mission critical software often, you have the development version, the release version, and the LTS. The LTS is older but has been tested to hell and back and everyone knows it works and works well. In regards to NHS certifying software, this is the one which should be the one it certifies. Yes, compared to the current release build it is dated, lacking features, but everyone, the NHS and those who choose to use it know it works and work well. Over time a trusted release build would get certified again annually or six-monthly and become the new Certified Build (not calling it LTS as support wise it doesn’t fit this example).  

 

Libre 

First off, I hate the Libre. Always have and no matter what people say about it probably always will. You are probably asking why shortly followed by the usual spiel about it being innovative, but I have to ask where is it innovative? I love innovation and the Libre has none. The whole it’s “innovative” thing really winds me up. So, firstly I’m going to explain why I dislike it so much. 

It’s a none real-time CGM. And before people start whining that it’s not a CGM, it is. End of, it has the same sensor technology as all the rest, the only difference is it doesn’t transmit that information real-time. That’s THE ONLY difference. You can call it Flash, or anything else, I’ll call it NFC (Near-Field Communication) because that’s what it. Some call this an innovation, but is it? Simple answer to that is no. It is actually a work around to bypass other problems with the design, mostly that on the power usage of the device. If it used a proper wireless system like Bluetooth it would need a larger battery as the chipset available at the time (no longer though) would result in more power being necessary and a larger device and also a higher price tag. This would produce a far better device as it would make the Libre a real-time CGM like those from Dexcom and Medtronic, but due to the power requirements necessary a design decision was made not to have Bluetooth in the current version (as said above about why). For those of you following the development of the Libre II, it has Bluetooth and as such the NFC chip and aerial have been removed from the design. The lack of Bluetooth (or ZigBee or any other wireless system) has resulted in the Libre being unable to do alarms. This is not some “innovative” idea to make it less intrusive or anything else, this was a decision forced onto it due to power requirements and as such in the Libre II alarms are a major “innovative” feature being added.  

Another major issue I have with it has been how it’s been released. Basically unfinished. The initial Libre when they came out have test points still on the mainboards, not something normally found unless it was a test device (which it was). Since release, the mainboard has been swapped, the aerial design has also changed, the sensor has been modified, and they’ve tried three different glues (at last count) on the pads. The supply of the device has also been massively limited which raises some large questions. The device is pretty much a beta device released early and being modified as it goes. Production limitations are aiding that for being done in smaller runs so modifications can be made to it as production can be scaled up later once the final design (the Libre II is worked out). As well as the limitations, there has been a lot of people experiencing failing sensors and failing glues even after the numour changes so far. All round, the Libre smells as pre-release testing which took off rather too well so the company went with it as best they could when in reality it was intended as an early in the wild test bed for the true CGM the Libre II. 

Why was it chosen first? Was it because it was “innovative”? Surely not as it’s dated technology lacking true innovations. So why? Was it because of their blinding marketing campaign paying off bloggers with trips to lovely places which got them wittering word for word marketing spiel and making the masses believe it? In doing so resulting in the masses sending endless streams of I want letters to decision makers? Highly likely. 

The Libre II is going to be a very interesting device. Still stuck to a set time use for a sensor which could prove interesting when it’s compared to the Dexcom G6. Both of these will have integration into various APS systems coming out “soon” so it’ll be very interesting how they work out cost wise. But as mentioned above, on cost when you go outside of the certifications the Dexcom will probably have it so it’ll all depend on that as to how the bean counters measure things. Whats said on paper or what occurs in reality.