So I've now been using the Dexcom for three weeks (my first sensor lasted 16 days, and I'm doing a second just to make sure it wasn't a fluke). One of the things I've been wondering is how those of you who use CGMs use the data. For example, if it says I'm low then I test and treat. But if it says I'm high, I've learned that often if I wait it'll start going down on its own and if I correct I end up going low. Or, if I'm going down, it doesn't necessarily mean I have to eat, because I might level off or start going up again.
I have managed to "head off" about two lows so far. But at other times I eat and end up going high. And I haven't managed to deal with highs any differently than I would when testing.
I know part of this is probably that my pump settings are off. I'm still trying to sort out my overnight basal rates! My overnight blood sugars are better than they were, but I'm still dropping or rising by about 3 mmol/L during the night. The daytime is even more chaotic. I haven't even tried working on those settings yet becasue I'm still trying to figure out why there are some days where I have lows and no highs and other days where I run high all day and other days where there's a mixture of both!
I do know for sure that when this sensor is done I will be buying the Dexcom G4 system.
I use the CGM data to tweak basal rates and learn from any patterns that emerge. Most CGM users watch the real-time data stream and adjust their behavior accordingly. That is a legitimate and useful feature.
A minority of CGMS users do what's called "retrospective analysis" which is just a fancy name for uploading the data and using report software, like Studio and Diasend, to spot patterns. I upload to Studio every day and to Diasend about once per week. I monitor my time in range, time low, standard deviation, and BG average for the 14-day, one month, and one quarter periods. I find that just following this data influences my daily behavior to positively affect these measures.
I believe that "time in range" is the most important diabetes control measure, by far. I see it as more important than A1c because, unlike the A1c metric, it does not hide high and low excursions.
I use the CGM to make changes every week. I think that the ability to change treatment as diabetes changes is a large strength of the CGM technology.
Congrats on your decision to buy the Dex G4 system. I believe it's well worth it.
Do you know where I can get the "time in range" information on Diasend? I don't have access to the Dexcom software (unless it can be downloaded)?
According to my data so far, my average BG is 7.5 (135) with a standard deviation of 2.6 (46) and min/max values of 2.2-17.9 (39-322). I have a target range of 4.0-7.0 (72-126) in Diasend, and so far 51% of readings are within that range, 44% are high, and 5% are low. If I set my upper limit target to 10 (180), which is what it is on my Dexcom, then only 15% of readings are high.
This is actually fantastic as far as I'm concerned! I'll be very curious what my next A1c (in a month) will be.
I've only had this for three weeks and already feel as if it's life-changing. I feel more empowered and much safer with diabetes than I have in a long time!
You will find that the dexcom changes your life. I have one since July and would not ever be without it. It don’t have a computer that is compatible with the dexcom but can get on one once o month just to see patterns. But it helps me every day , last night I treated for dinner like I always would but I guess I did not have as much carbs as I thought at 3 am it woke me I was at 60 and was able to stop a low . I do find that high reads are most of the time not as high as it indicates . When I get I high reading I test and find it to be off a good amount. Having said that I let it help me hour by hour and I adjust as the day moves along. I don’t have a pump so it’s a bit different then you but it’s changed my life .
I can’t believe you got 16 days … Wonderful. I got 14 one week but most of the time just get 7 days . Which is fine with me
Good luck . Your a1c will defently make you happy after 3 months with the dexcom.
It looks like you've found the time in range number. It's on the compilation page and is shown as a pie chart for CGM. The other pie chart is for the meter readings.
The Dexcom Studio software can be downloaded from the website (US). It looks like you can download Studio from the Dex Canada website, too. Then you'll see the more familiar mmol/L units.
I set my target range to 65-140 mg/dl (3.6-7.8 mmol/L). You seem like a person for which this technology will work well. You'll need an infra-red (IR) adaptor for your pump and most meters to get that data into Diasend as well. Ask Animas about the IR adaptor. I think they sent me one, upon request, for free. I use a Roche Accu-Chek Aviva meter. The IR adaptors may be available from them, too.
Now that you're aware of your baseline numbers, you'll be aware of any progress. Good luck with continued success.
AUC means area under the curve. It’s used a lot in medical research as a standard for comparative purposes. I was never formally trained in statistics so I don’t yet grasp the full meaning.
Here’s what I understand. Area, in math terms, is the product of length and width. In this context, the length is the x-axis or time. The width is the y-axis, or BG value. The calculus integral function is the way that the actual area under the curve is calculated.
In our BG context, the smaller the area under the curve, the better. Obviously, even a perfect BG level will still have some area under the curve. So the nominal value is still some positive number. I’m thinking that the high AUC is the extent to which your numbers exceed your high limit. Conversely your low numbers are the extent to which your numbers fall below your low threshold.
Given all that, I still don’t know how the number that Diasend calculates for AUC is figured. I’ve done wikipedia and other searches but still don’t get AUC 100%. Good question. Maybe someone else can answer both of us.
Bear with me, I drew this in paint. Terrys right, so they are calculating the amount of area under the curve that fits into each range: low (blue), normal(green), high (red). All the area adds up to one, so you might get 0.4, 0.2, and 0.4. They are probably defining it so that you want a large percent in the normal range, here that might be 20%. You will want low AUC in the high and low ranges. But, someone would confirm this. Things kinda depend of how they choose define things. This is, traditionally, how AUC works.
So what's the differnece between this and the % spent high and low? My AUC for high was 0.4 and the % was 44%, while low AUC was 0 and 5%. So evidently the AUC (at least in Diasend) does not count anything in the ones, which seems less useful. Is there ever a time when AUC and % might differ?
I think I was wrong because I found this:Pharmokinetics definition of AUC I'll call them tomorrow and ask. I think their closed for the night. I was wondering about this, myself, when I saw that AUC measurement. I can't find it in the Dexcom reports, now. Still looking...I bet this is how they measure IOB. Are you pulling the AUC measurement out of your pump or the Dexcom or Diasend?
Well this is even more confusing. Why is your high AUC 1 but you only have 8% of readings high, while my AUC is 0.4 but I have 16% of readings high? Maybe what you really want is a higher number, not lower...?
My understanding is that AUC relative to BG, is a negative number to represent under the low end of your range, and positive number for 'amount' above the high limit. On the MM pump w/sensor data, AUC is calculated and shown on the pump screen.
On Dexcom, I don't see it labeled as AUC, but seems to display on the Patterns screen in Studio, at the bottom. You can check Dexcom Studio guide for more info. However, it appears they don't provide a numeric number, like Minimed, that could be tracked over time.
Good question. Someone here knows this answer. I've been interested in the AUC for some time but always felt my understanding was incomplete or incorrect.
There are 2 numbers, representing the 'area' above your high range, and different number for below your low range. If 8% of time you were above by exactly 5 pts, it would be a lower number than 8% of the time below by exactly 30 pts.
Based on percent in range, you would see 8% above, 8% low, so it would not distinguish the 'magnitude'.
The goal for both AUC high and low is to be as close to 0 as possible !
If you can use Diasend to reset your high/low range, you can probably see how that effects the AUC result.
Relative to MM's suggestion, I changed my target range to match Jen's, 72 mg/dl (4 mmol/L) to 180 mg/dl (10 mmol/L). My high AUC went from 1 to 0 and my low AUC went from 0 to 1. Yet I had 32 values above 180. Is that just due to rounding?
I'm still missing a few pieces of this puzzle. On top of that I can hear my grade school math teacher telling me that area is always expressed in "square units," like square inches, square yards, square millimeters. I've never seen AUC expressed in anything but just a number. And Diasend seems to just use whole numbers, no decimal points.
I agree that MM's idea that the closer AUC is to zero, the better. Whenever I've had stellar numbers, the AUC has been zero.
I've never heard before that Medtronic CGM-enabled pumps express AUC on their screen. Perhaps we can look up the users' manual for the requisite pumps and see what their narrative has to say about AUC.
I don't know, but its something to do with how they calculate IOB, right? This guy gets to it at min 5 and he has a better accent and cartoons. Better Explanation of AUC I can't find a calculator, but I'd bet their using the logarithmic algorithm...ln(0) = 1?? I think I'm learning things that I have always wanted to know. This is a great post.