Poll - Dosing from CGM data alone

I’ve been wondering about this for some time and I’d love to see the results as a whole, within our community, for those that make dosing decisions based solely off of the CGM data versus those that do finger sticks to verify bgs before dosing.

A yes vote is CGM only dosing decisions and a No vote is finger stick verification before dosing.

Of course this poll is only relevant is you use a CGM.

  • Yes (I dose off of CGM data alone)
  • No (I don’t dose off of CGM data alone)

0 voters


I do both, depending on situation and reading on dexcom. I still use G4, so this may change if using G5 or G6.

Before looking at dexcom, i usually have a good idea of what it might be. That also guides whether i do bg check.

So i didn’t make a choice on the poll.


Like @MM2, sometimes I doublecheck with a fingerstick, sometimes not.


I might double-check with a fingerstick if our son is running very high and I’m about to give a huge bolus. Or if his blood sugar isn’t doing what I expect based on the insulin and food he’s been given.


The poll isn’t asking about exceptions… Only the normal bolus method that you use most of the time. There are always exceptions that can and will be treated differently but this poll is asking your NORMAL way of doing things. If you have no normal and do differently every time then I would personally put that as a vote for not Bolusing based on the CGM number alone.

I answered yes, but we do have a specific method which makes it hard to reply to your question as asked:

  • We dose based on CGM

  • We take sugar based on CGM unless we are below 75. Below 75, we always fingerstick before taking carbs for a low AFTER the first time we have taken carbs


So you answered correctly according to how I’m asking the question. If I understand what you’re saying, your bullet #1 is your “normal” bolus method and your bullet #2 is the exception for you.

1 Like

I have a G4 and I don’t find it reliably accurate – it’s often off by 1, 2 or sometimes even 3 mmol (20 to 50 mg), despite calibrations – and so I don’t like to dose from that number UNLESS I’m eating out (in which case, because of unknown timings and guessed-at carb counts, it can be a crapshoot anyway). Also the CGM has a mealtime lag that is often more than 15 minutes, and if a meal extends over a couple of hours or more, I’ll fingerstick in advance of each course because the CGM won’t have caught up with the last carbs I consumed. If I had a G5 or G6 I think I’d be much more comfortable dosing from its numbers.


@Beacher, I am not sure there is a difference between the G4 and the G5 in terms of the quality of numbers. There definitely is one with the G6 though—I am waiting to see if it works better for us. I am cheered that several experienced members have had positive things to say about G6 accuracy for them.


Depends on whether running the improved G4 algorithm on the receiving device or not. I forget - maybe it was/is called the SW505?

If running that algorithm (SW505) and with the sensors for G4/G5 identical so the results you would expect (for better or worse) are exactly the same between G4 and G5 systems.

In terms of G6, whether better or worse, it is a different algorithm and a different sensor. So the results are not the same. With no regard for the results being better or worse as compared to the G4/G5.


I don’t think I’ve commented on my success with the G6 yet, so I’ll say here that I find it very reliable and dose off of it all the time.

Also, contrary to some of the suppositions that are floating around here, I have a normal range BMI, and I exercise hard the the point of exhaustion three to five times a week, 15 to 25K total every week. So it apparently can work well for lean exercisers.

The only way I suspect I may be different than some other posters, is that I always wear it on my abdomen, and that is virgin real estate for me since I have never pumped or used a CGM before, and have rarely injected in my abdomen (using it only when stuck sitting up on a plane etc.)


I actually answered “no” based on my experience with the G5. But with recently starting on the G6 I have been fingersticking much less often. When I do, the Dexcom is within 1-3 mg/dl of the result.

So I guess the G6 is good to dose from.


Really good to read that!

Same with my son. He is lean although not exceedingly so, and his CGM works very well (his sensor duration is another issue…). My son has a somewhat comparable exercise profile as yours. He typically runs 5K per day, most days in the week.

I think Dexcom works well for many lean users. But it seems that the set of those for whom it does not work well includes an unexpectedly large percentage of lean athletes. So it is an issue of odds, really, I think.

That shows the G6 as really excellent for you, I am impressed! We see this kind of precision on the G5 (within 5mg/dl) only for about 40% (+/-) of the sensors.


I didn’t answer the poll because I do both. It depends on the situation. I would say that if my CGM (G5) has been pretty on track I use it for dosing if my BG is fairly on target and stable. If my BG is on the low end or high end, or if it is going up or down quickly, then I will check with a fingerstick before I dose. I also do a fingerstick after exercise if I am having a snack.


I personally think that there is no ‘Normal’ with T1D. It’s all about exceptions.


I don’t know. According to the votes there are plenty of people who bolus one way or the other.

If your exceptions are more frequent than not then I would add that as a no vote.

1 Like

As the poll seems to be getting sticky around 2/3 (Yes) - 1/3 (No), at least in my experience this makes sense. I think the sensors work well for many, but not all. The key for the manufacturer is going to be how do they increase the performance in the group they don’t work so well in. I would love to see the manufacturer’s put out some literature on the subject.


And can tend toward unreliable for… the… not as lean exercisers. :roll_eyes:

I exercise a lot, have for years, and can definitely get to being slender, but I don’t exactly fit the “can’t pinch an inch” crowd. I got inches. Enough not to be a problem pinching them.


They know there are some issues, but I don’t think they know what causes some groups (i.e some pediatrics) to have issues.

According to this podcast talking specifically about G6 with the Dexcom CEO they are looking into.

Starts at 13:38.


Like many here, I also do both. If I feel like the cgm is behaving I will dose from it. First Am I never do as cgm is always off 20 to 30 pts, so I fingerstick. If I feel off somehow I verify cgm with a fingerstick before dosing