The NHS England Hybrid Closed Loop experiment comes a step closer to implementation

A post from Partha Kar on X piqued my, and many other people’s, interest this week.

It revealed the list of technologies for which commercial terms had been agreed with NHS England, and would be made available to users via clinics.

What was perhaps most interesting was how the data was presented, and therefore how we as users should interpret it.

We were shown the following page from an official document:

• Which of the items on the list will be shipped together as a system?
• Why is Medtronic Simplera Sync not on the list?
• Where is Libre2 Plus?
• Why is there a Medtrum system on the list?

This has created a number of questions.

Taking these in turn…

Which of the items on the list will be shipped together as a system?

By presenting a mix and match list of components and including the statement:

ICB and trust leads are encouraged to contact NHS Supply Chain and engage with suppliers to find out what combinations, at the cost-effective price, work best for their local populations.

This has led to questions on X about what will be available in each region, given the reticence some ICBs have shown over CGM, and whether multiple CGM options with specific pumps or algorithms will be made available, or whether end users will be limited to a single CGM option.

I’d like to be able to say that all will be available everywhere over time, especially with the funding model in place, but I imagine that some ICBs will make that much harder than it should be.

Why is Medtronic Simplera Sync not on the list?

This is most likely to be down to timing. Whilst the sensor may have been seen all over Diabetes Social Media, it’s not yet launched in the UK. I’d expect this to come in to play as Medtronic update their platform.

Where is Libre2 Plus?

That was answered previously.

The Libre2 Plus is already on the Drug Tariff, making it prescribable by GPs. And it costs the same as Libre2 per day.

So for anyone using a pump that can be turned into a Hybrid Closed Loop by adding a Libre2 Plus, they can simply get their prescription changed. It provides a cheap and simple path to HCL.

Why is there a Medtrum system on the list?

Perhaps the biggest question on this information release.

Of all the options open to NHS England, the Medtrum Nano A8 system is the only one in which neither the algorithm, nor any other of the components have been reviewed and approved by the FDA.

It is notable that the Medtrum system is not discussed in any literature about the NHS study. We know the Omnipod 5 wasn’t either, but that wasn’t available at the time. Medtrum were advertising their system in other jurisdictions when it ran.

NHS England have also highlighted it separately on their document, as APGO, suggesting its a separate algorithm. Even prior to this being marketed, Medtrum had talked about predictive low glucose suspend in their Easypatch app, and in the UK that still isn’t offered.

What’s also notable about the Medtrum system is that there is no published data available as to its performance.

The CGM, as listed in the document, references the A8 Nano. As does the patch pump. I assume these are what I’ve used previously and recently talked about.

If this is the case then it raises questions as to what everything is licensed for. What follows is an interpretation of English language, rather than a legal opinion. As an end user, it’s how I’d interpret the information.

MHRA license

Firstly, the MHRA license that Medtrum currently has doesn’t contain any references to algorithmic control of insulin delivery. Their license is of type “Ambulatory insulin infusion administration unit/remote control system”.

According to the Global Medical Device Nomenclature (GMDN), this is what that means.

Tandem and Medtronic use a different description that clearly states that an algorithm is involved in the control of insulin delivery.

Meanwhile, insulet have another version, which, this time, refers to Hybrid Closed Loop.

The Medtrum license was granted in 2021, so it’s not outside the realms of possibility that they were planning something with a Hybrid Closed Loop, and that by obtaining this license they might avoid additional licensing for an algorithmic solution while they didnt have one available.

Personally, I’d have expected a different license, especially as Tandem, Insulet and Medtronic had all registered their products multiple months prior to the Medtrum license being granted.

CamDiab, whose product I consider is most similar to Medtrum APGO (a separate AID algorithm on a phone/device), and which NHS England seems to class similarly, has a different license altogether, again registered prior to Medtrum’s.

It’s worth noting that they are all classified as class IIb medical devices, which also includes dumb insulin pumps.

CE Mark

Medtrum user guides contain a reference to a TUV Rheinland CE Mark. This appears to relate to having the appropriate ISO, MDD and MDSAP standards in place for the design, development, manufacturing and distribution of CGM systems and Insulin Management Systems. I may have missed another specific CE Mark for APGO though.

CGM accuracy and classification

The CGM data for the A8 Nano suggests that it doesn’t meet the iCGM standard. A Medtrum document found on-line at the Czech State Institute for Drug Control suggests that the ±15%/15mg/dl value is 78.4%  The iCGM requirement is that below 70mg/dl this needs to be more than 85% while between 70mg/dl and 180mg/dl, this only needs to exceed 70%.

With a value of 78.4% across the entire range, it seems questionable statistically whether the sensor meets that requirement. Historically, Medtrum have also tended to measure ±20%/20mg/dl values below 100mg/dl in their literature so it is also not clear whether the data relates to an iCGM-style standard, or something else.

What does this all mean?

I think there are a number of things that come out of this.

• There is zero public performance data available for the Medtrum HCL system. There was supposed to be a trial in Europe starting in the early part of this year, but there’s nothing I can find at all.
• A lot of the “information” about what Medtrum is offering is unclear. While I know I legally have no clue about MHRA licensing terms, the plain English reading  doesn’t suggest a Hybrid Closed Loop is being described.
• Based on the data I’ve seen (which isn’t really published, but seems to be publicly available with an internet search), there are questions about the performance of the CGM at low glucose levels, and if this is the same sensor I’ve used, I wouldn’t want to run an HCL off it.

I’ve asked a couple of healthcsre professionals if they have any data from Medtrum, and the result was a blank. They’ve not seen any data either.

For me, the question is, “What data was shown to NHS England?”, and why can’t we see it?

Are we in a good place for HCL roll out in England?

I think we are. I suspect the challenges about system selection and choice will be overcome over time, especially with the funding model being used.

But I also think ICBs need to be very careful about what they choose to offer to users.

Any clinician helping users to choose a system, and anyone paying for it, should be incredibly wary of offering a system that doesn’t publicly provide pivotal trial performance data.

Would you buy a car where the manufacturer told you it would do 50 miles to the gallon and 0-60 in 2.8 seconds without anything to back up that claim? Then why would anyone choose a Hybrid Closed Loop because “of course it works”?