Revisiting the Dexcom G6, now that #WeAreNotWaiting and have adapted to it…

Back at the end of March, when Dexcom launched the G6, I questioned how this might affect those using WeAreNotWaiting open source Diabetes technology. It raised a number of questions. These were:

• Can the G6 be used with the existing tools?
• Can it be restarted?
• What was the cost of this new product going to be?
• How long can a single sensor be made to last?
• Could we use it for looping offline?
• Was the ongoing battery replacement process that many use for the G5 viable?
• How well does it work?

A number of us have been working with the G6 now for (in some cases) months, and there are answers available for these questions.

1. Can the G6 be used with the existing tools?

This is a resounding “YES”. Thanks to the work from Katie DiSimone and “JamorHam”, both Loop and xDrip+ can be used with the Dexcom G6. And they can be used in full non-calibration mode. Obviously a great win for the WeAreNotWaiting community. Support is also on the way for Spike.

It turned out that the differences from the G5 were not as great as feared and in the protocol being used and these are now available.

2. Can it be restarted?

Again, a resounding yes. The first publication of restarts was again by Katie DiSimone over at her blog. It turns out that there are four ways you can do it if using the official app, and that Spike can be used to do a transmitter reset as well. All great news to the ears of G6 users. But wait, there’s more.

If you’re an xDrip+ user, then you can choose to set up as a G6, and it will undertake the pre-emptive start, 9 days after starting the transmitter. This restarts the transmitter, doesn’t seem to lose data and seems to allow you to continue in non-calibration mode. Obviously, all good stuff, and to do this is very straightforward.

To set up the G6, you’ll need to do the following, and then start as normal in Settings. When you start a G6 it also asks for the sensor ID to allow the on-G6 non-calibration functions. This is written on the removable backing paper for the adhesive, so before applying the G6, especially if you intend to soak the sensor, take a photo of that paper.

1. Long press the blood drop the configuration menu will appear
2. Select “Source Wizard Button” on this menu and touch the background to hide the menu
3. Select the “Start Source Wizard Setup” button
4. Select G5/G6

Following the instructions from here will allow the optimal set up for the G6 and set up all that’s required for restarting the sensor in the background.

3. What is the cost of this going to be?

In the UK, the cost of individual items is roughly the same as the G5 was, however, the subscription based model works out as a better deal. I’m not going to call it cheap, because it’s not, however, it works out less expensive for a full 12 months of 24×7 coverage. What does this mean?

It means that the price of the subscription model is now £159 per month for 4 transmitters and 1 sensor every ten days, where the previous equivalent coverage (4 transmitters and 1 sensor every 7 days) cost more and the half coverage (1 sensor per 14 days) cost £160 per month. Overall then, you’re getting full coverage at the same cost as the half coverage model on the G5 plan.

Given that we’ve answered the question about restarting already, you’d have to find out how long you could make a sensor last for to determine what model worked best for you financially.

4. How long can a single sensor be made to last?

One thing I and others have noted about the G6 is that with an adhesive change, we’re struggling to get the sensors to remain stuck down for the full ten days. I’ve commented in Facebook groups that I think Grifgrips are now essential, whereas I used to be able to wear a G4/G5 for 10-12 days without additional sticking.

I’ve only used one sensor so far, and I needed to apply a Grifgrip at about six days. The second sensor I applied it after three. The sensor died at 15 days, having stopped giving readings. That’s with SkinTac and Grifgrip applied. For comparison, at 18 days with the G5 I used to stop using it due to noisy data. We’ll see how the second one manages, but anecdotal data from groups on Social Media suggest that many people are not seeing longevity in the G6 sensors.

It was noticeable that when I removed the sensor, the adhesive pad on the G6 had no stickiness left at all. Potentially something worth bearing in mind when looking at the costs.

5. Could we use it for looping offline?

Very easy to answer this one. Yes for Loop, AndroidAPS and OpenAPS. All require either the modified version of Loop or xDrip+ in a recently nightly version (start from 15th July).

6. Can you replace the transmitter battery?

Once again, yes you can, and Katie has written it up here. It’s worth noting that she mentions that it’s more difficult than the G5, but it works.

This also shows that the battery in use is the CR1632, compared to the 2x SR1130 in the G5. Looking at this from a capacity perspective, the CR1632 has a capacity of 140mAh at 3V while each SR1130W has 85mAh at 1.55V. Given that the G5 batteries are connected serially to generate the necessary voltage, the G6 is carrying significantly more capacity, so it’s possible it will last longer than the G5, dependent on what the transmitter algorithm is doing (amended following Richard Oppenheimer comment).

This raises the question as to whether the G6 transmitter will last as long as the G5. This will be dependent on power draw. Maybe the G6 uses less power? We’ll have to wait and see until someone has run one to its full life.

7. How well does it work?

This is the big question. And it seems there are varying answers. If we assume that the majority of people for whom it works well don’t post on social media, then I think it probably works very well.

Speaking for me personally, I’ve found it to be very good. Over the life of my first sensor, I had very little drift from bloods when used without calibration, and I was happy to loop from it. What I mean by this is that I attached it, calibrated it via xDrip+ after a couple of hours, then left it. The next calibration required was on the day it was restarted. Compared to the G5, this is hugely better. I’m aware that there are a lot of people who don’t find the need to calibrate the G5. I wasn’t one of those. I’d find it drifted to >1mmol/l out after roughly 24-36 hours, so needed to calibrate daily to ensure accuracy. I’ll run a full fingerprick compared dataset at some point in the future, but am awaiting a couple of other CGMs before that happens.

As I’ve already mentioned, the adhesive on the patch wasn’t so good, and I’m not the only person to observe this. I’ve heard of a fair few people that have had issues with it sticking properly. It will be interesting to see how this one plays out.

The other observation for me (and possibly much more major if you’re not the one wearing the device) is the transmitter range. Used with xDrip+, it’s not great (and seems worse than the G5) and we continue to hear from users on the official apps that they find its range to be poor. Even to the extent that sleeping with a phone on the bedside table, they are losing signal. I’m not sure what Dexcom can do to improve this, so it may remain a challenge for many.

In conclusion

The Dexcom G6, to me at least, appears to be well on the way to working with all the WeAreNotWaiting options, and my fears previously have proven to be unfounded. In and of itself, the G6 system appears to present some challenges to some users, and it’s still not what you could consider a cheap option. Hopefully the updated way of using it, its lack of need of calibration and longer life will enable healthcare providers to pick it up and accept the benefits.

In the meantime, we are still not waiting, and for those who want CGM, but can’t afford to pay full price or need a broader set of options, G6 does work with these tools.