Poll - Dosing from CGM data alone



I really don’t dose off my CGM… anything more than very small, insignificant boluses. This is an example of the things I’ve seen that contribute to my thinking… My CGM was showing a fast rising trend, and having had TWO bananas after my run tonight, it confirmed what I was afraid would happen— a big rebound. I almost “took care of it”, which would’ve meant a fairly aggressive bolus (a true rebound off of 2 bananas could mean I was headed for the 300s). Almost. I decided to double check just to make sure… and saw the 117. Decided against the couple of units.

This isn’t to say other people have this problem, and there are plenty of times my sensor is just fine, but these kinds of instances just make me leery to bolus without double checking first.

Oh! And I should include in here that it actually was a rising blood sugar. I watched it for a while after this just to try to get a handle on what was happening, and my blood sugar did continue to climb to about a 140, but my sensor at that point was showing a ~185 with two arrows up. I’m glad I didn’t do the insulin… I was dropping again within the next 45 minutes.


If I think the G6 is wrong I draw blood.

The G6 gives me a trend, I bolus off the trend not the number.


I really like the G6, and I do understand the importance of using it for the trend and not the value. There’s plenty of times throughout the day that works out well. That’s why I put the picture in the thread… Tonight I actually thought the G6 was confirming what I had been anticipating… which was why I was tempted to just do the insulin. Turns out both the CGM and my instinct were overshooting what was happening.


Sorry, that wasn’t meant as a criticism; I replied off your post after responding to the original (eight day old) poll. The result was utterly out of context.

I do do broadly what you do; I only use the G6 as a trend, I only trust the reading if it is stable and is in the range 100-150 and not always even then. Otherwise I use all the other stuff I’ve developed over the years - how much I’ve eaten, how my muscles feel, whether I’m annoying the people around me and what the BG meter says.


The fact you include that in your “measures” just shows you are a much more considerate person than I am. :grin:

And I didn’t take it as a criticism at all. I respond to everyone like that. Even when talking about sunshine and picnics. :grin: How do you like your G6? And have you used any others?


Combo answer. I would dose off of cgm data when I’d been fasting, sometimes, if it appears to be giving meaningful consistent data… but never in a million years would I do it for corrections.


How do you like your G6? And have you used any others?

I gave up on an earlier model (the G4?); it woke me up repeatedly and failed after about 24 hours. They didn’t refund my money so I blacklisted Dexcom for a few years. The G6 is good, but remembering my previous experience still recreates a bad taste in my mouth so I won’t say more than that.
I’ve dropped calibrating the thing; it seemed to produce bad effects when done outside the 100-150 mg/dl range. I also just run it until it comes unstuck - xDrip+ has a “preemptive restart” option that avoids the PITA two hour wait period and I’ve yet to see any issues with running at least 20 days. Indeed, the first few days of a G6 seem less stable than the 20+ days, but that’s just a feeling.
It’s difficult to see how it could work in a closed loop without some mods. I’d trust it at 100+ with no significant up/down, but to use that closed loop means maintaining BG at 100+; I don’t want to do that.
I suspect eventually they will retune the CGMs to be accurate at 80-120 and require a closed loop.