Excellent.
we typically see a rise in blood sugar within 10 to 15 minutes of our son eating. It’s usually registered on Dexcom at about the same time, maybe 5 to 10 minutes behind that. The faster the rise, the longer the lag, so if he’s spiking from 100 to 300 in 30 minutes, Dexcom may be 50 to 100 points off for the first 30 minutes.
Also, Dexcom isn’t just getting a simple reading from interstitial glucose, it’s also using an algorithm to try to map that interstitial reading onto a BG reading. I don’t fully understand their algorithm but my impression is that a large part of it is not just denoising, but trying to remove some of that lag.
1 Like
I am jealous. After a meal, my G6 reading slowly, very slowly climbs over two hours. I can go from 100 at mealtime to the mid- to upper- 200’s two hours later. If it goes over 300 I take a bolus of fast acting insulin to bring it back down. My goal is to go to bed with a G6 reading around 150. I normally wake up with a G6 reading near 100. And the movie repeats.
But, how does a sensor failure present in your case?
jumpy data (i.e. “triples” where the dots seem to follow a trend and then all of a sudden one jumps up and then gets back “in line”). Large periods where it says “sensor error” or “no readings,” or even “replace sensor now,” or “check BG” or “recalibrate in 15 minutes.” Basically, actual errors.
Our errors are almost always ones where we get no data, not bad data. The algorithm is meant to be pretty conservative. it usually gives no data rather than give a number it’s uncertain or confused about it, from what I understand.
2 Likes
I can see why that would be frustrating. Is he an active kid and possibly bumping the sensor? Where is the sensor?
Let’s hope the sensors become or are becoming more consistent, I’m UK based, but my latest batch of sensors have been superb so far. I had issues and failures initially with the G6, in addition to sensors now happily lasting the ten days, my last two sensors have lasted 20 days with nigh on perfect accuracy; I removed them at 20 days while they were still working perfectly, so they would have gone longer. Bizarrely they have also been relatively accurate on the first day so maybe they have sent me a batch of super sensors by mistake 
OR the latest batch also came with a new transmitter, something else to think about?
On the note of lag, I usually have very little, it tracks my BG within minutes and sometimes, very nearly parallels any rises or falls, lags occasionally but seems like magic to me. As I’m low carb, I tend to run between 4.4 and 6 mmol (80 - 110 mgdl) and a “high” for me would be 7 mmol (125 mgdl), so I don’t tend to have very quick drops or rises on any sort of regular basis…
I think the G6 has the potential to be pretty outstanding and here is hoping it becomes a lot more consistent and realises that potential. Dexcom UK customer service has been brill so far, I’m hoping it stays that way and sends some of its pleasant competence over to the USA; I certainly don’t want the unpleasant incompetence heading in the opposite direction 
1 Like
I have been wondering this also.
We have been not having great luck with G6 longevity and most sensors have lasted us 8 or 9 days. Sometimes they run the full 10 days and even a couple times we were able to extend to 12 or 13 days.
The last two sensors, I jumped out of my normal stock (use oldest sensor from shelf first) and used the newest sensor we had which has expiration date June 7, 2020. Both of these sensors have done surprisingly well.
Maybe we got lucky last two sensors or maybe Dexcom has made a minor improvement in the sensors.
1 Like
Really? And? Have you any thoughts??
1 Like
That doesn’t stop people from buying their CGM supplies from Walgreens. There is no reason that they have to buy from Dexcom directly.
Maybe I am lucky that every time I reach Tech Support, I reach a U.S. resident. I start a pleasant conversation with them and unless they are remarkably well-trained in American minutiae and extremely good at emulating an American accent, they are not Filipino. Maybe Tandem users have a different experience- I don’t have a pump, so I wouldn’t know about the Tandem support.
I started using the G6 shortly after it was available. I had more than a few sensor problems in the first few months, but as with any new product there are always early production problems. My last request for a sensor replacement was more than six months ago. By the way, if you were a former employee, that last line could be a legal risk.
1 Like
I think that was already addressed previously.
How did you get Control-IQ so soon—were you in the trial? I have lots of questions about Control-IQ, but let me ask just two. How well does it moderate BG spikes if you underestimate your carb intake or if you get a hormone-induced spike from stress? How sensitive is it to errors in your basal settings?
168 people took part in the clinical trial.
I would love to hear some first hand experience and get insights before the Control-IQ algorithm is actually approved and released !!!
1 Like
Do you have any newer sensors?
I would be extremely interested if you notice significant improvements on newer sensors.
I can send you some if you want.
Knowing these limitations with replacements, I would, quite honestly just lie when I talk to them. It’s a ridiculous argument and as such I would do whatever I had to do to get a replacement for the item that I paid for - even if that means lie to the customer service agents. ::shrug::
We don’t have problems with Edgepark customer service. The only issue we have had is their expectation of “full payment upfront” before they ship sensors. The company we used before Edgepark would send sensors and send us the bill afterward. So we have to now make sure we always have enough funds in HSA when we’re making that phone call.
Me, neither. I haven’t been on it for seven years, or even close, but I’ve watched it like a hawk since I have been. That 30-60 minute estimate is very familiar… just not for the G6. That’s that Medtronic lag 
, @SteveMann. I know all about that kind of lag, really I do, but never for the G6. If you’re genuinely seeing that, maybe you should try a different CGM?
Our next sensor (same box so same expiration June 7, 2020) has also hit 10 full days and we are now doing a restart on this sensor.
This is the first box of G6 sensors since we switched from G5 to G6 (actually a year ago in Sept '18) that we have even hit 10 days (plus bonus restart days) on each of the three sensors in the box.
I really think something has recently changed with the G6 sensors. For the better.
4 Likes
The current D6 batch is definitely the worst yet in my 2 years using dexcom.
This example is from Day 2. The low values are all 20-30 too low.
Dexcom continue to change the values retrospectively but it is just getting worse. There are no missing values - dexcom are actually changing them. How does the 30% rule work with this I wonder 
Reading 76 (drop -24), FingerStick 109
2 Likes
What is the expiration date of the G6 sensors you are talking about?
This has been in place since the release of the G6. This is graph smoothing that Dexcom has implemented with the G6.
As soon as the current cgm data point comes in then graph smoothing is applied to the previous value which has just now become 5 minutes old and is now historical. Graph smoothing is only applied once and is not reapplied to that data point.
The 30/30 rule (G6) is applicable to the current value for which graph smoothing has not yet been applied.