DexCom Data Analysis

[Cyborg]

I started this new thread to post comparison data from the DexCom CGMS vs. blood glucose readings taken from the OneTouch Ultra meter.

Let the fun begin!

7:00 Sync Meter with Device

Time DexCom Ultra Error
4/12/06 7:37 PM 97 107 9.35%
4/12/06 7:50 PM 101 100 1.00%

[spike]

I started this new thread to post comparison data from the DexCom CGMS vs. blood glucose readings taken from the OneTouch Ultra meter.

Let the fun begin!

7:00 Sync Meter with Device

Time DexCom Ultra Error
4/12/06 7:37 PM 97 107 9.35%
4/12/06 7:50 PM 101 100 1.00%


Lookin' good! Let's see how it does in hypo territory and hypers, and how consistent it is. I also would like to see a correlation of numbers, adjusted for the time lag of eight minutes. Isn't that the known time lag of the sampling method, compared to finger sticks?

Do you have fast moving bg's? I can rise 100 points in less than 30 minutes and drop at the same rate, or slightly faster. I'm looking forward to seeing how much of a lag there is with rapidly changing bg's.

[Cyborg]

I also would like to see a correlation of numbers, adjusted for the time lag of eight minutes. Isn't that the known time lag of the sampling method, compared to finger sticks?

Yes, I think the rep was quoting 7 or 8 minutes. That's why I'm taking back to back readings. If there are any discrepencies in the values, I don't want lag being used as an excuse.

Do you have fast moving bg's? I can rise 100 points in less than 30 minutes and drop at the same rate, or slightly faster. I'm looking forward to seeing how much of a lag there is with rapidly changing bg's.

My liver dump in the morning is pretty strong, but it's controlled by basal rate changes so I may not show a huge rise as when it wasn't controlled. I may go ahead and fast tomorrow till lunch and check my basals.

[Cyborg]

My fingers hurt!

7:00 Sync Meter with Device

Time DexCom Ultra
4/12/06 7:37 PM 97 107
4/12/06 7:50 PM 101 100
4/12/06 8:42 PM 85 82
4/12/06 8:54 PM 88 87
4/12/06 10:12 PM 87 101
4/12/06 10:22 PM 112 119
4/12/06 11:03 PM 173 154
4/12/06 11:18 PM 170 153
4/12/06 11:47 PM 174 154
4/12/06 11:57 PM 179 163

[Shotokan]

My fingers hurt!

7:00 Sync Meter with Device

Time DexCom Ultra
4/12/06 7:37 PM 97 107
4/12/06 7:50 PM 101 100
4/12/06 8:42 PM 85 82
4/12/06 8:54 PM 88 87
4/12/06 10:12 PM 87 101
4/12/06 10:22 PM 112 119
4/12/06 11:03 PM 173 154
4/12/06 11:18 PM 170 153
4/12/06 11:47 PM 174 154
4/12/06 11:57 PM 179 163

Mine hurt just thinking about it! Do you think the agreement would differ depending on what your BG is when you sync it? You synced it when your BG was around 100. I wonder how different things would be if you synced it when your BG is 150.

With all this interest in your "experiment" do you feel a little like a lab rat?

[Cyborg]

I'm not sure about the calibration since I did calibrate several times today before the final one at 7:00 pm. I just calibrated again and the device has resynced. I'm around 160 right now and I'm done for the night. So I guess I'll answer your question better in the morning...

And yes, I feel like I'm in a clinical trial where I pay for everything! I figure it this way, this information will be useful to many interested in CGMS technology. Also, if the device does not do what it says, it is under warranty. ;)

[spike]

My fingers hurt!

7:00 Sync Meter with Device

Time DexCom Ultra
4/12/06 7:37 PM 97 107
4/12/06 7:50 PM 101 100
4/12/06 8:42 PM 85 82
4/12/06 8:54 PM 88 87
4/12/06 10:12 PM 87 101
4/12/06 10:22 PM 112 119
4/12/06 11:03 PM 173 154
4/12/06 11:18 PM 170 153
4/12/06 11:47 PM 174 154
4/12/06 11:57 PM 179 163


I know this is asking a lot, but do you have any other meters with strips that you could double check the results with? I primarily use an US, but I also use a Compact and Flash. From my testing, I feel that: The Compact goes high, compared to the US, at around 230+. The Flash is a bit high compared to the other two, at hypo levels. Those deviations aren't consistent, so that erratic behavior, compared to the US, makes me leery of depending on the Compact at 200+ and the Flash at <75. My US keeps me safe. <g>

[seacomp]

4/12/06 10:12 PM 87 101
4/12/06 10:22 PM 112 119
4/12/06 11:03 PM 173 154

My sympathy for your fingers.:help:
But, I wonder if you do a bit more, not testing, reporting. I.E., what happened after 10:12 to cause the spike?
That is, add in eating and other relevant events. Thanks.

[Cyborg]

I don't have any other meters except the UltraSmarts. I wanted to use the Ultra meter because I have test strips for that meter, I can use it to sync the DexCom device, it matches my UltraSmart exactly and is used by the DexCom people. If I had another meter and strips, I would endure the extra finger sticks for the better cause.

I will mark up the data more to help. For the data last night, I ate dinner at approximately 9 PM. I ate or did nothing else the entire evening except post on the forum and do a little work on the pc. The spike in bg was the delay in digestion from the Symlin. Remember, the "real" numbers are from the meter, not the CGMS unit, so the spike wasn't so big.

[Cyborg]

Time.......................DexCom......Ultra
4/13/06 4:23 AM.......49.............70
4/13/06 4:25 AM Sync Meter with Device
4/13/06 6:59 AM.......110............92
4/13/06 7:11 AM.......115............90
4/13/06 8:00 AM.......114............109
4/13/06 8:09 AM.......109............96
4/13/06 8:11 AM Sync Meter with Device
4/13/06 8:41 AM.......121............100
4/13/06 8:51 AM.......119............104
12 Carb Breakfast @ 9:00 AM
Here I started taking the CGMS reading at the same time as the bg meter and 10 minutes afterwards (when I remembered)
4/13/06 10:57 AM......84/78.........78
4/13/06 1:17 PM.......78/82..........77
13 Carb Lunch @ 1:30 PM
4/13/06 3:27 PM.......141/143.......116
4/13/06 4:18 PM.......161.............124 - Received High Alarm
4/13/06 5:30 PM Sync Meter with Device
4/13/06 7:31 PM.......76/77..........79

[JediSkipdogg]

Wow, those numbers are pretty off in some places. It appears easy to tell when a sync is going bad. I'm just amazed these things can't keep sync for a long time. A 141 and a 116 are suppose to be the same? If I didn't double check on that 141 and I bolused for it to bring me back down to 100 and I was really running 116, I could go near hypo. ANd I know when kids get their hands on this and even young adults, they WILL make corrections WITHOUT checking. This seems like a good help, but ONLY if used properly and without much trust.

[spike]

Wow, those numbers are pretty off in some places. It appears easy to tell when a sync is going bad. I'm just amazed these things can't keep sync for a long time. A 141 and a 116 are suppose to be the same? If I didn't double check on that 141 and I bolused for it to bring me back down to 100 and I was really running 116, I could go near hypo. ANd I know when kids get their hands on this and even young adults, they WILL make corrections WITHOUT checking. This seems like a good help, but ONLY if used properly and without much trust.

Pray tell what good is any metering system if you can't rely nearly 100% on it's readings? So far, my US is the ONLY meter I trust.

[JediSkipdogg]

Pray tell what good is any metering system if you can't rely nearly 100% on it's readings? So far, my US is the ONLY meter I trust.

Hmmmm, go to www.myspace.com and join the 10 or so diabetes forums over there. You will see what I mean. Kids leaving pump infusion sets in for 1-2 weeks at a time. They make correction factors based on what they feel without testing. They drive while they are 50 and figure they can make it home 20 minutes instead of stopping at a Shell station for food. Do I need to continue?

A mature diabetic needs to trust his meter. An immature diabetic just needs something to give him a number +/- 25%. I haven't used the Flash yet, although I would like to try, but I think I'm sticking to Lifescan meters only.

[spike]

Hmmmm, go to www.myspace.com and join the 10 or so diabetes forums over there. You will see what I mean. Kids leaving pump infusion sets in for 1-2 weeks at a time. They make correction factors based on what they feel without testing. They drive while they are 50 and figure they can make it home 20 minutes instead of stopping at a Shell station for food. Do I need to continue?

A mature diabetic needs to trust his meter. An immature diabetic just needs something to give him a number +/- 25%. I haven't used the Flash yet, although I would like to try, but I think I'm sticking to Lifescan meters only.

I didn't think we were arguing, but from the way you phrased the first paragraph, I'm wondering if we are on the same page. I'm not disputing anything you wrote, but it seems irrelevant to the comment of mine that you quoted. I'm confused as to what point you are making, relative to my previous post.

Dave

[JediSkipdogg]

I didn't think we were arguing, but from the way you phrased the first paragraph, I'm wondering if we are on the same page. I'm not disputing anything you wrote, but it seems irrelevant to the comment of mine that you quoted. I'm confused as to what point you are making, relative to my previous post.

Dave

I thought you were using some sarcasm towards me. But I got it now. My bad :shakehand

[spike]

I thought you were using some sarcasm towards me. But I got it now. My bad :shakehand

I wasn't being sarcastic at all. I might be on occasion, but not in the post you are referring to. No problemo! :)

[Cyborg]

I spoke with DexCom about the discrepancies today and we decided to keep the sensor in and continue to see if the device syncs and stays synced. To make a long story short, we never got to do the initial calibration on this sensor and because of that, it may be the cause of these differences in bg values. I'm still collecting data on this sensor and will continue to do so. DexCom has already informed me that they will replace 2 sensors free of charge. I am trying to do everything possible to help them along with this device.

[JediSkipdogg]

I'm glad you are continuing on and willing to help them instead of just sending it back. I'm also glad to hear that Dexcom is willing to work with you as well. It's a two way support structure like that that will improve the product.

What do you mean you didn't do the initial calibration? What is the difference between what you did and what you are suppose to do?

[Cyborg]

What do you mean you didn't do the initial calibration? What is the difference between what you did and what you are suppose to do?

I didn't have an Ultra meter 2 hours after inserting the sensor when the 2 blood droplet screen appeared, which indicates time for the new sensor calibration. I hit the clear button and the alert and screen would go to sleep for 1/2 hour then alert again. This continued until I finally got an Ultra to use for calibration. When we went to do this initial calibration of the sensor, the device was no longer requesting an initial calibration (2 blood droplet screen), but was only requesting a normal calibration (1 blood droplet screen). We proceeded rather than changing out the sensor since they had never seen this happen before. So now there is the possibility that this is the cause of the discrepancies in the readings.

[JediSkipdogg]

What is this two blood droplet screen you keep referring to? Do you mean you have to do two seperate fingersticks for the initial calibration? If so, how close do the two readings have to be since we all know readings can vary for multiple reasons.

[Cyborg]

What is this two blood droplet screen you keep referring to? Do you mean you have to do two seperate fingersticks for the initial calibration? If so, how close do the two readings have to be since we all know readings can vary for multiple reasons.

It's just the "initial calibration screen", as opposed to a single droplet which is the "normal calibration screen". The initial calibration is done once for a new sensor. Normal calibrations are done every 12 hours, usually.

[Cyborg]

Time....................DexCom....Ultra
4/13/06 5:30 PM Sync Meter with Device
4/13/06 7:31 PM....76/77.........79
4/13/06 7:45 PM Sync Meter with Device
78 Carb Dinner @ 8:00 PM
4/13/06 10:00 PM...130...........140
4/13/06 10:54 PM...153/139.....124
4/14/06 12:12 AM...117...........133
4/14/06 12:45 AM...148...........137
4/14/06 1:00 AM Sync Meter with Device
4/14/06 2:03 AM.....LOW..........114 Got Low Alarm (Under 55)
4/14/06 2:05 AM Sync Meter with Device
4/14/06 7:33 AM.....129/130 ....118
4/14/06 8:46 AM.....159...........130
12 Carb Breakfast @ 8:50 AM
4/14/06 11:39 AM....98............99
4/14/06 12:49 PM....70............89 Got Low Alarm (Under 70)
4/14/06 12:49 PM Sync Meter with Device
4/14/06 1:12 PM......54............? Got Low Alarm (Under 55)
Ok, I give up… No more verification on this sensor!

[Cyborg]

Well, we started out pretty well. Until about 15 minutes ago I was pretty impressed:

http://www.diabetesforums.com/gallery/files/2/6/5/9/4-19-06.jpg

[Cyborg]

lol... Now it says I'm 194! I'm really 96 still...

:banghead:

[Shotokan]

Well, it was doing great during the first 6 or 7 hours. Nuts! I really wanted this thing to work!

[Cyborg]

I got woken up at 4 times last nite before I took the receiver and put it in my office. Then the darned thing still somehow picked up a false reading and an alarm went off and woke me again. I was pushing buttons trying to turn the thing off but couldn't figure out how (lucky I couldn't find a hammer, as it was about 3 in the morning). The alarms were all alarms indicating my bg was about 250 to 300 or under 70, all the while in reality my sugar was near normal. :mad:

Ok, so I'm tired. Maybe not the best time to post...

[gettingby]

Oh man. Sorry Cyborg. Hopefully this cycle won't continue but if it does, I would definitely be getting my money back and telling these people to check out the "bugs" again. This thing seems to be infested with them. LOL (sorry, too early for humor).

[JediSkipdogg]

I agree I would contact the company and another thing if you are really up to it is to send a letter to the FDA. Normally I am not for people complaining to the FDA, however, it appears that this device is not up to par at all and in the wrong hands could be dangerous and deadly.

[gettingby]

I agree I would contact the company and another thing if you are really up to it is to send a letter to the FDA. Normally I am not for people complaining to the FDA, however, it appears that this device is not up to par at all and in the wrong hands could be dangerous and deadly.
I have to agree here.
This device + being in the wrong hands = Possible disaster.:(
I do applaud you for stepping up to the plate and making the effort to stick with it, even through the rough times. :congrats:

[Cyborg]

I must be a glutton for punishment. I called technical support and got some suggestions and decided to leave the sensor in. I'll be posting more data later tonight. I'm running out of test strips, so the data comparisons will be reduced until I can get more.

[Shotokan]

Did they come up with a possible reason for the large disagreements? What did they suggest?

[Cyborg]

Apparently it is in the acceptable range according to Dexcom. Even 30-40 points. I don't agree as this was not how the accuracy of the device was represented. I'm more concerned about the number of false alarms at night. I have not had a solid nights sleep with the device yet. The only suggestion I got today was to do a calibration at dinner time rather than before bedtime. :stupid:

[Shotokan]

Well, if it makes you feel any better you probably saw the other guy's post about the Guardian. He shipped that back as well. Let's hope the Navigator does a better job!

[JediSkipdogg]

Apparently it is in the acceptable range according to Dexcom. Even 30-40 points. I don't agree as this was not how the accuracy of the device was represented. I'm more concerned about the number of false alarms at night. I have not had a solid nights sleep with the device yet. The only suggestion I got today was to do a calibration at dinner time rather than before bedtime. :stupid:

30-40 points is deadly. If someone is running 80 and is driving a car and the CGMS reads 120, they may make a correction to 90 (if that is what they like to run.) Well, 120-90 = 30, but in reality, 80-30 = 50........BLACKOUT possibly. Enough side.

[Cyborg]

Here's some fresh data. I did get woken up with 2 false alarms last night. It seems the unit has more difficulty at night when I need it most.

http://www.diabetesforums.com/gallery/files/2/6/5/9/4-20-06.jpg

[spike]

Apparently it is in the acceptable range according to Dexcom. Even 30-40 points. I don't agree as this was not how the accuracy of the device was represented. I'm more concerned about the number of false alarms at night. I have not had a solid nights sleep with the device yet. The only suggestion I got today was to do a calibration at dinner time rather than before bedtime. :stupid:

40 points "acceptable"?? Not to THIS fingersticking guy!

[Shotokan]

Here's some fresh data. I did get woken up with 2 false alarms last night. It seems the unit has more difficulty at night when I need it most.

http://www.diabetesforums.com/gallery/files/2/6/5/9/4-20-06.jpg

That's not too bad, except for the waking-up part. Maybe you should set the alarm threshold a little lower.

[seacomp]

30-40 points is deadly. If someone is running 80 and is driving a car and the CGMS reads 120, they may make a correction to 90 (if that is what they like to run.) Well, 120-90 = 30, but in reality, 80-30 = 50........BLACKOUT possibly. Enough side.
I wonder if the company has people like us in mind. By and large the people in the Forum are working as hard as they can to keep their BG at target all or most of the time. The regulars here, at least, are into tight control in one way or the other. To them, for us, 30 or 40 points is a big deal for exactly the reason you point out.
Many, if not most, diabetics (*sorry Deus), are happy to know their BG is at 200, not 400.
Maybe, like with an automobile you need a "stick" shift :driver: if you want the best performance?

[Shotokan]

I would be relatively happy with a 10 percent error so that it would measure +/- 5 points when my BG is 50 and +/- 25 points when my BG is at 250. But I would not be happy with +/- 25 points when my BG is 75.

[Lynne1]

I was wondering the same thing about my meter. I'm using the Paradigm link and getting hypo feeling at 75-85. Could it be that my meter is off? I remember testing with the paradigm link vs my accucheck advantage and the paradigm link being 25 pts higher...most of the time.:help: What to do?

[Shotokan]

A lot of people have complained here and elsewhere about the accuracy of the Paradigm link meter. Many people have abandoned it altogether. The problem might lie with either the strips or with the meter itself. But from a practical standpoint it doesn't matter if you can't trust the results.

[spike]

I was wondering the same thing about my meter. I'm using the Paradigm link and getting hypo feeling at 75-85. Could it be that my meter is off? I remember testing with the paradigm link vs my accucheck advantage and the paradigm link being 25 pts higher...most of the time.:help: What to do?

I absolutely won't use the BD Logic meter due to it's erratic readings. In a just world, they would all be returned to the mfgr and sent off to a landfill. My LifeScan meters are my "go to" meters, and I'm sorta ok with the Compact, but it reads high, above about 230 or so, compared to other meters. The Flash reads too high for me in the hypo range. How many times have I mentioned that.

[Dewey]

Here's some fresh data. I did get woken up with 2 false alarms last night. It seems the unit has more difficulty at night when I need it most.

http://www.diabetesforums.com/gallery/files/2/6/5/9/4-20-06.jpg
I'm sorry that the system is giving such trouble, Cyborg. :( Like Cin said, I admire that you've stuck with it to help others out. I sure hope that Dexcom gets these issues fixed & that the system turns out to be even better in the long run.

Truth be told, as I've learned more over the years, I no longer fully trust any system. I recently had an incident where my Freestyle Flash read 411! :eek: I started to correct, but wanted to ensure that the reading wasn't off, so re-tested. The second reading said 177, so I tested on the One Touch UltraSmart & got a 168. I stopped my bolus about mid-way, but kept watch to ensure I didn't go too low.

It seems that with the technology available today, there'd be a good way to make an accurate, continuous system that has minimal failures or false readings. I sure hope we all see that day soon.

[Cyborg]

Thanks... I was really hoping the device would work as represented. When I was told that 30-40 point variances from actual readings was acceptable while I was close to my target, that was when I realized the device was not ready for me. When the thing is detecting 40 points higher than my actual blood glucose, that means I could be around 30 before I ever get my custom alarm of 70! Someone (not Dexcom) suggested lowering my custom alarm from 70 to stop the false alarms, but that would only make that scenario even worse). I've not gotten a single nights sleep without false high or low alarms.

At this point, I've asked for a refund. They assured me I could get a refund before I started using the second kit they sent me. Now it seems they are delaying the refund "decision". I'm not sure what is going to happen. I have removed all sensors and everything is on hold at this point in time. If I do not get my refund, I will be forced to contact my bank to get my money back and they will get a bad reputation, which I'm sure they do not want. It's sad that it's come down to this as I feel I've worked closely with them to provide accurate feedback. Unfortunately, I also feel that I've been burned... :(

[spike]

Thanks... I was really hoping the device would work as represented. When I was told that 30-40 point variances from actual readings was acceptable while I was close to my target, that was when I realized the device was not ready for me. When the thing is detecting 40 points higher than my actual blood glucose, that means I could be around 30 before I ever get my custom alarm of 70! Someone (not Dexcom) suggested lowering my custom alarm from 70 to stop the false alarms, but that would only make that scenario even worse). I've not gotten a single nights sleep without false high or low alarms.

At this point, I've asked for a refund. They assured me I could get a refund before I started using the second kit they sent me. Now it seems they are delaying the refund "decision". I'm not sure what is going to happen. I have removed all sensors and everything is on hold at this point in time. If I do not get my refund, I will be forced to contact my bank to get my money back and they will get a bad reputation, which I'm sure they do not want. It's sad that it's come down to this as I feel I've worked closely with them to provide accurate feedback. Unfortunately, I also feel that I've been burned... :(

Is the refund promise in writing? MM has already told me on 2 occasions there will be no refunds on their CGMS. Don't know if they'll stick to that, but that's their official position (at least with respect to the convesations I had with them. other employees may give other responses)

[Cyborg]

Is the refund promise in writing? MM has already told me on 2 occasions there will be no refunds on their CGMS. Don't know if they'll stick to that, but that's their official position (at least with respect to the convesations I had with them. other employees may give other responses)

I don't have anything in writing, just their word. I'll soon know how good their word really is...

[gettingby]

I don't have anything in writing, just their word. I'll soon know how good their word really is...
Don't back down and don't let them use you as a doormat !!!!