More DexCom: Addicted to the Line

It’s been just over two weeks, and I’m hooked. I can’t imagine going back to simply sticking my finger several times a day, or every few hours at most…

As the DexCom rep put it, it’s like we used to be driving along in a car with the windshield blocked out, and only every now and then would someone pull the cover off so we could have a quick glimpse of where we were on the Diabetes Road (not even a look at the road ahead, just a quick peak at where we were that very moment). Then suddenly we got this new navigation system, the Continuous Glucose Monitor, which lifted the cover off the windshield permanently, so we could actually see — all the time — where we are, where we’re going, and even where we’ve been. Wow!

That said, I’ve had some serious accuracy issues with the (first-generation) DexCom. Last weekend, I had an afternoon where the DexCom had me running 180-190, while the OneTouch (traditional meter necessary for calibrating the DexCom) had me in the 140-150 range. That’s a big difference to me — the difference between correcting and “riding out the high,” actually. That evening, DexCom had me at 156, while the OneTouch clocked in at 228. Wtf? Whaddoo I do now? Double-check with my old standard FreeStyle Flash meter, of course. 216. OK, correction city!

The following morning, DexCom made me smile with a reading of 128. But I lost my grin when I checked the OneTouch and got 188. The FreeStyle agreed, so I dialed up an aggressive breakfast dose.

By this point, understandably, my faith in the whole thing was going out the window. So I called DexCom tech support. They’re very responsive and helpful, I must say. But that doesn’t make the device more accurate. The DexCom support folks told me that up to a 30% delta between the CGM and the traditional meter was documented at this point. Aaargh!


I know this accuracy gap is quite upsetting to some people. But I firmly believe that one cannot and should not give up too soon in this early, pioneering phase of CGM use. Hear me out…

As all of us in the current Field Study are realizing: right now it’s not so much about the individual number(s) as it is about spotting the trends.

Despite gaps between the CGM and traditional meter, the DexCom still gives me the line, which means a lot. I’m constantly reminded that no matter what number I get with a fingerstick, I never have any idea which direction I’m going. Enter the DexCom, which gives me a line graph that quite accurately reflects whether I’m having an upswing, downswing, or a nice steady lull.

And luckily, my newest sensor seems to be more spot-on, so my smile’s coming back. I even went swimming with it yesterday and it kept on tickin’ right under the shower patch!

So what has my line graph done for me so far? (pictured above is an example only) A lot, actually. It’s showed me where my “problem phases” lie. My post-meal results after breakfast and lunch consistently sucked. So I’m taking action: tofu and scrambled eggs for breakfast; earlier, more aggressive injections (no more waiting till I’ve cleared at least half my plate); and I’ve tightened my insulin-to-carb ratio a bit. And I can see — in real-time even — that my numbers are evening out!

The other thing to remember here is that a continuous monitor doesn’t actually do anything to improve your glucose control; rather, it gives you valuable information about what you can do for better D-management. I’ll let you know how it goes…


13 Responses

  1. Kevin
    Kevin July 10, 2006 at 8:06 am | | Reply

    Thanks for the update, Amy.

    So I didn’t realize DexCom had their software available yet (or did I and I’ve just forgotten?). How do you like it?

  2. AmyT
    AmyT July 10, 2006 at 11:09 am | | Reply

    Ah, the software is available for physicians now — so you have to go in to have your stuff downloaded. The patient software is due out by the end of the year, I understand.

  3. Kassie
    Kassie July 10, 2006 at 11:49 am | | Reply

    Amy, are you finding that the on-screen graph gives you sufficient info, or are you hankerin’ for the home version of the software?

  4. Nick
    Nick July 10, 2006 at 3:23 pm | | Reply

    Amy, you are right that the value of this device is in displaying your blood sugar’s rate-of-change; that is, how quickly it’s rising or falling. How well does it do this?

    The device’s accuracy is also an issue. You’ve indicated some readings are 30 percent higher than your glucometer’s readings.

    Are the reading ***consistently*** 30 percent higher? Is the
    error always 30 percent, whether your blood sugar’s low, normal, or high?

    If the error is consistent, then you can correct for it. If it’s not consistent, then the device displays inconsistent behavior. By carefully measuring its error for a range of different blood sugars values, say from 50 to 250 mg/dl, you ought to be able to ascertain whether or not the device’s error is consistent across the entire range.

  5. AmyT
    AmyT July 10, 2006 at 3:29 pm | | Reply

    Let’s see: error rate is not consistent, but can be up to 30%. As noted, some sensors perform better than others. Common in First Gen tech products.

    And Kassie: so far I’m happy with getting the printouts periodically, as long as I have the device telling me what’s going on 24/7. It really is amazing how I’ve been able to steer myself into the target range!

  6. Nick
    Nick July 10, 2006 at 7:07 pm | | Reply

    Why are there four different curves, of four different colors, in the graph?

  7. AmyT
    AmyT July 11, 2006 at 8:38 am | | Reply

    Hey Guys,
    This graph is just an example, not my personal data. The different colors represent four separate days.

    I can see that I’d better use more specific data next time :)

  8. susan fisher
    susan fisher July 11, 2006 at 9:09 am | | Reply

    Hey Amy – we met at the field study ;)

    My last sensor after the first day when the device was ‘learning’ my curve, was never more then 10 pts off. My current sensor was about 60 pts off last night, so I definitely think it varies. They say the accuracy of the device will improve with the more fingersticks you feed it – and that first sensor, I was testing like mad because I didn’t trust it yet.

    The curve or trend is very very cool. It tells me that I’m plateauing after a lunch bolus (oops, need more!), or that I am climbing too quickly after a meal (oops, didn’t give enough!).

    It also tells me that I am utterly stable and am safe to drive! Previously if I tested 100, well, I didn’t know if I was rising or falling, and I don’t feel my lows, so I’d eat before driving. Now if my fingerstick says 100 and my graph is flat or rising, I’m good to go. Same goes with feeling safe to sleep…

    I love it.

  9. ksc
    ksc July 12, 2006 at 7:21 am | | Reply

    Have you had an example of catching a rising high or sinking low despite the #’s being slightly off? That is something I’d love to have available if I ever get a sensor.

  10. christmasx2
    christmasx2 July 13, 2006 at 6:58 pm | | Reply

    Thanks so much for posting all of this great info! I look for updates almost daily, please keep them coming.

    I am thinking seriously about making the investment in a DEXCOM CGM, but am tentative because I wasted a lot of money on the glucowatch (what a disaster that was). At this point, I have decided to wait until the Fall to see what shakes out between Dexcom and the new Minimed device.

    Again, thanks for all of the info!

  11. j. Turner
    j. Turner July 18, 2006 at 11:30 am | | Reply

    The main point to remember is that the Dexcom is not a replacement for fingerstick glucose testing. It’s purpose is to show you the OVERALL pattern of your glucose readings, how it fluctuates during the day, how you respond to certain foods or activities, etc. It’s the whole picture of your control, not each indivdual stick. None of the CGMS are designed for this. And in most cases, the difference in glucose between fingerstick and CGMS are not usually enough to facilitate a change in therapy.

  12. Sam
    Sam July 21, 2006 at 9:18 am | | Reply

    I have had very similar experiences Amy and have found myself checking MORE now that I use an STS than before I did. One thing that helps accuracy is to do a bunch of finger stick tests early on and callibrate every single one. That said, to the questioners out there, the accuracy is never incredibl reliable… the problem is that it is reliable enough at times that you start to trust it! Another big issue I am having is that I see high blood sugar readings and give myself insulin. Then I wait 20 minutes and I’m still high (of course it’s this way! I’m just not used to seeing myself high that long). So I treat again. Then look out bellllloooooooww. Some wicked lows result. And Dex can’t keep up so it will say 140 and I’ll be at 65. Anyone experiencing this as well? One last point – I think the Navigator is going to be the gold standard – the Guardian has been on the market for years and done nothing and led to lots of frustrations – Abbott has shown better accuracy data in trials than either Dex or MiniMed.

  13. Ken Whistler
    Ken Whistler August 6, 2007 at 11:51 pm | | Reply

    Diabetic Complications are the permanent changes that occur after prediabetes stage. Body indicates these changes by steady symptoms like pain in the eyes, etc. Diabetic complications may damage body organs like heart, kidney, eye, feet, skin.

Leave a Reply