Dexcom 8Gxxxx restart-block and pre-soak

Thought I’d do a separate thread on this since it has come up in a couple of discussions. As we know, the new Dexcom G6 transmitters, with SN’s starting with 8G, are designed to prevent sensor restarts. They do this by checking for “insertion trauma”: if the BG track is too steady during the 2hr warm-up period, the new algorithm deems it an attempted a restart and invalidates the sensor. The concern this raises is, what about doing a pre-soak before initiating a legitimately new sensor? The point of pre-soaking (inserting the new sensor while the old one is still running, and giving it 12-24 hrs before swapping over the transmitter) is precisely to allow it to acclimate and thus ameliorate the erratic readings we often see for the first 12-24 hours of a new session. So there’s been a question whether doing a presoak would result in the system invalidating the sensor because it doesn’t see the “insertion trauma” it’s looking for.

I’ve now done two pre-soak sensor insertions on 8G transmitter.
First one (ten days ago):
Pre-soak, 12 hrs.
Result: no invalidation
Sensor Behavior: not very effective—I had erratic readings for the first 12 hrs, then it settled down.

Second one (today):
Pre-soak, 16 hrs
Result: No invalidation.
Sensor Behavior: Starting off kinda flakey. How do you get an initial post-warm up data point with a sharp down arrow? I mean, down compared to what? It was also 25 pts high compared to a finger stick, and so what the hey, I calibrated. Came down, but now it’s trending back up again.***

So… we’ll see how the next several hours go. When I’ve done pre-soak before, it has seemed more effective than this. Could be just random, could be there’s something about the new algorithm that is making this technique less effective than it was. If I get the same erratic readings for the next 12hrs that I got last time, I may not bother doing it again. Since it does seem safe to try, I’d like to see what results anyone else gets.

***ETA: as it turns out, it actually settled down in less than an hour and has been spot on ever since.


Geez, the more I read about the latest G6 trickery employed by Dexcom, the more I want to stick with the G5. A simple 1-day soaking of a G5 gives me excellent results when I start it up after roughly 24 hours. I’ve tried soaking for 1/2 a day a number of times and that just doesn’t give me very good results, but a full day’s soaking and I’m always good for 7 days of great readings. I don’t want any flakiness on any day of a session, be it a 7 day, 10 day, or in the future, 14 day session. No thanks.

1 Like

To follow up from the above, which I wrote only about 1/2 an hour after the warm up, the sensor settled down within the hour and has been accurate and stable since.

Bottom line: I think we can go ahead and do the presoak and not worry about it getting invalidated as a sensor restart. As for sensor restarts themselves, that seems to be a no-go for now. If someone’s got a hack that works, I haven’t seen it yet.


Thanks for the update and the trial, very helpful !

Well, “trickery”—I mean, it does feel a bit like a game of cat-and-mouse. We’ve been taking advantage of the loopholes they’ve left open, but it’s within their rights (and responsibilities, from a regulatory perspective) to close them. So from their p.o.v. it’s more like “due diligence.” Not that I have to like it.

YDMV: even in spite of this, I wouldn’t choose to go back to my G5; there are just so many things I prefer about the G6, leading with the accuracy. From what I’d read, I thought it was going to be marginally better, maybe indistinguishably so, but instead in my case it has been quite dramatic, and that has had cascading effects for my whole management. My A1C is running in the 5’s for the first time since I was dx’d, I’m hardly using finger sticks anymore, the no-calibration thing turns out to be much more of a convenience than I’d anticipated, etc etc.

But I agree that the no-restart thing is a PITA, not least because I relied on it to make sure of having an extra sensor or three in reserve for emergencies. It bugs me not having that peace of mind. A 14-day sensor won’t change that either, if there are no restarts, b/c you’ll still be prevented from using extended sessions to build up a reserve.

I wish I’d known about the change sooner or I might have held on to my old transmitter and tried doing that battery-replacement thing. But I tossed it. Sigh.

It’s nice to hear that you really are getting increased accuracy. That is something to look forward to for sure. Coming from Enlites I’ve been happy with the vast improvement in G5 accuracy but there are times when it tends to run lower by 10-30 points with stable bg’s so if the G6 is more accurate overall I’d like that. Because I don’t have to pay for them the lack of restarts wouldn’t dampen my enthusiasm, but I feel badly for those paying OOP if they are no longer able to restart.

The first year or so I was doing G5 restarts, also to build up a “stash”. Since the stash in now adequate, I just change every 7 days, since I rarely could go more than 9-10 days on the best performing sensors, anyway.

I’ve never thrown out a transmitter. I mark them with magic marker and keep them as “dummies” for doing the sensor-soaks. And just in case I get put onto the G6 by Medicare, don’t like them, are unable to switch back to G5, then I’d try my hand and battery replacement/reset transmitter and use up more of my G5’s.

Did you have so many problems with erratic readings while using older models of the Dexcom? What was the last model you used? G5? I’ve been on G4 for 11 years until recently, so I’m out of the loop on newer sensor models.

G5 is also known to be erratic the first 12-24 hrs and I experienced that to varying degrees, sensor to sensor. When I first switched to the G6 it didn’t seem any more pronounced than the G5, but more recently it does seem so. Obviously the non-calibration thing entails changes in the algorithm vs the older versions, especially this business of compensating for “insertion trauma.” May explain why “compression lows” are more of a thing with the G6, especially in that initial period. Like I said above, this latest sensor started out with a pretty cross-eyed result, but as it turned out this was very short-lived, like one or two excursions that cleared within the first hour and it has been super-accurate since then.

My last one, also on the 8G transmitter, took more like 6hrs to smooth out, but I hadn’t given it as long a presoak, a little shy of 12 hrs vs 16 hrs for the current one. I’ve seen others report good results with even shorter ones, and sometimes that worked for me too, but I think those were all pre-8G. So 16 seems to be the sweet spot for me. I’ll probably stick with that for now.

One additional factor: when I talked to a Dx tech support guy about whether the new algorithm would invalidate due to pre-soaking, we discussed that initial erraticism and he said “You know, it’s ok to do a calibration when it’s like that.” So on this latest sensor I did go ahead and do one after the initial weird reading, and that may have been a factor in why it settled down so quickly.

Good tip, but please let us know if in the future, it was more of a one-off occurrence of increased accuracy or seems to be a consistent method of getting good first-day numbers.

I’m using the 8G now, my readings for the first 8-10 hours have been the worst I’ve seen with a Dexcom, including the G5. I replaced my sensor at 2am, the thing has been 30-50 points off. My pump has been suspending delivery because the sensor has continually said my BG is below 80, for the last few hours it’s reading 50-40, when by BG was 80.

1 Like

That’s why I won’t let a sensor dictate my pump’s insulin delivery. I don’t care the brand, or the promises, I know better than to trust the tech. At some point I’d love to see someone make a “minority report” version of a sensor. 3 sensors in one, with the odd one “out” and the other two averaged to hopefully obtain more accurate CGM data. Even then I’m not sure I’d trust it, but it might be a step towards getting me to trust it after lots of feedback from other users on accuracy. :slight_smile:

1 Like

Yeesh. Both of my post-8G sensors have been much better than that, though the most recent one more so than the first. Do you presoak them? Not that you really should have to, mind, but it does seem to help in my case.

1 Like

No, I don’t presoak. Not because I think it’s a bad idea but that I’m horribly lazy with my sensor changes.


It’s frustrating. I never had this issue before Dexcom started screwing around with the transmitters and sensors. Even my first 24 hours were very accurate, now they suck. With my pump suspending delivery on false readings it just compounds the problem.

1 Like

Yeah, I hear that. I’m the same way, mostly, just adds another layer onto the finagling nature of things. But I’ve been a little more consistent over the last few changes.

1 Like

I’m definitely trying it out next sensor change. Thanks for the advice on how to do it with the best results.

1 Like

I think its amazing/strange/crazy how different MY experience with G6 is, compared to the majority of you on here.
I’ve been on G6 for just about a year now and experienced terrible signal loss & readings for probably the first 8 months. The bluetooth signal seemed to compete with my phone’s bluetooth (I use the Dexcom receiver as a preference because I also have an older phone and battery & storage space is extremely limited) I “upgraded” directly from G4, which I’d been on for years.
But SOMETHING changed recently and I’ve been able to restart my G6 sensors. Sometimes only for a few days before the readings get too wonky and I get frustrated with them and give up. Just within the last month and a half I was actually able to restart a G6 sensor 3 times for almost a full months worth of readings. Amazed the heck out of me, but as you all are well aware, if I can do that at least once or twice a year, that’s a decent supply build up right there!

1 Like

Did you restart a sensor with 8G serial number? If so, please tell how?

1 Like

I’m pretty sure they did start with 8 but I don’t have any around to check. Waiting on my replacement order now.
I forget who the tudiabetes user is who I snagged this from but here is the method: when your sensor is done, don’t take it off, just hit start new sensor and choose “no code.” Finish the procedure to start the sensor. Then set an alarm for 15 minutes. After that 15 minutes, go in and cancel the sensor. It will ask if you’re sure, say yes. Then start it AGAIN but this time, enter a code you’ve used previously. I usually save the paper tabs when I take them off my sensors & use a random one. If you don’t have any old ones, check the others in your supply and use one of them.

Hope that helps. As i mentioned, I can usually get at least a few more days out of the sensor but sometimes more.

The earlier 80 and 81xxxx transmitters were allowing restart per your process. But not with latest 8Gxxxxx transmitter.