I was wondering about pumps that integrate with CGMs to deliver correction bolus insulin. I don’t have a pump but as a retired electronic technician I am curious how things work.
I’ve had several of these “No Readings” alert with the message that it might take up to 3 hours to restore. It usually happens around day 7 or 8 of sensor life (Dexcom G6). The first indication is a sudden fall of BG. Yesterday 5/15 (day 8) I was sitting at the table when BG dropped 18mg/dl in 10 minutes to No Reading alert. With the same sensor early this morning I got 2. The first was a 33mg/dl drop in 6 minutes. the second was 33mg/dl in 10 minutes. I am close to the receiver and phone and don’t sleep on my belly.
Anyway how comfortable are people with pumps that use CGM data? In my first 90 days of using the Dexcom G6 system I have leaned to take odd readings with a grain of salt and a finger stick or two. I over treated a faux low early on that I ended up out on my road in the middle of the night walking off a self induced hyper.
My starter kit contained 9 sensors all of the same lot number. I was suspecting them, but Dexcom sent me 3 of various lot numbers. Two were perfect for 10 days while the third is the one I am currently using.
I can’t stand it that I cannot figure out the problem. It is what I spent my whole life doing, diagnosing and correcting. I just don’t think I would be comfortable with the Dex G6 system controlling my insulin.
Your G6 seems to be working like mine. Can’t answer your question about Loop with unreliable CGM sensors as I have not tried that.
My first 6 months or so I never had G6 problems, in fact I was usually able to extend sensors for a few extra days. But lately I have rarely been able to get the full 10 days out of each sensor. Day 8/9 often (but not always) gives me those false rapid drops and sensor time outs. After a cumulative 3 hours of sensor time outs (or if the sensor totally dies) I change the sensor and send in the form to Dexcom for a free replacement.
Most reliable approach for me is a very tight tape job around the sensor, using skin tac for good adhesion to your skin. That has given me 10 days on some but not all G6 sensors lately. Even with that, I’ve had “pressure lows” when sleeping on the sensor that caused the false low and the sensor timeout.
@John58 Thanks, I am so glad it didn’t fail on day 7 as I did my first fairly long bicycle ride of 2hr19min. I was watching my BG on my watch. It was a great ride that I began with 9g nut bar and 20g of a glucose cocktail I created. We stop mid ride, my BG had slowly dropped from 140 to 95. I ate another bar and the rest of that delicious drink. BG rose up to 159 and then fell off to, I forget, about 110 at the end.
My group would’ve been irritated with me if I made them stop so I could finger stick.
I am very comfortable with my Loop system using Dexcom G6 CGM data to modulate my basal rates to control my blood glucose. Loop sets 30 minute temporary basal rates in response to CGM data and announced eating. If Loop loses the CGM data stream, it will continue the last 30-minute temp basal rate until it expires. Then it reverts to the background basal rate schedule programmed into the pump.
So, it’s important to keep the pump’s basal rate schedule and insulin sensitivity factors fresh. What I’m trying to say is that an accurate CGM data stream is important overall, temporary glitches are overcome with Loop’s reversion to the pump programmed basal schedule.
You do still need to monitor the accuracy of the CGM by doing regular fingersticks and when your symptoms don’t match CGM readings. In other words, this is not a set-it and forget-it system.
When I adopted Loop 4 1/2 years ago, I was surprised at how well it could control my glucose levels. It did a better job than I could and it’s level of attention is far superior to mine. I can’t speak to the quality of Tandem’s Control IQ or Medtronic’s hybrid systems but I’ve read many positive accounts from those using those systems, especially the Tandem one.
I would never willingly go back to using a manual pump system. I get much better results with less effort using Loop.
What happens to Control-IQ technology during and after a lost CGM signal?
Control-IQ technology will continue to adjust basal insulin delivery and deliver any required Automatic Correction Boluses for the first 15 minutes that the CGM transmitter and pump are out of range.
Once the Out of Range condition is present for 20 minutes, Control-IQ technology will not adjust insulin delivery. Instead, the pump will deliver the programmed basal rate in the active Personal Profile.
It is important to note that Control-IQ technology will remain ON but be inactive, with a grey diamond icon on the Home Screen, until the pump and transmitter reestablish communications. Once the pump and CGM are communicating again, Control-IQ technology will resume automatically.
Based on the trend you posted, the cgm outage occurs after 2 periods of dropping readings. Tandem pump would have adjusted basal down, so that when it hits the no cgm, would likely be at very low or zero basal.
Your example may be due to compression on sensor that leads to that pattern of cgm data having drops and gaps.
I think it depends on your body type. 95% of my G5 sensors made it through the 7 days without a drop, however, when I was upgraded to the 10 day G6 at least 75-80% start to drop readings on day 8. My drops tend to be a lot longer than yours as they can last up to 3 hours, especially on day 9. My drops are not compression lows.
So for me, I would never even consider controlling a pump until Dexcom is far more accurate and reduces the drops to next to nil. In the meantime, I am on MDI with a digital pen that microdoses down to 0.1 units of insulin which works perfectly for me. Although the G6 is by no means perfect, it sure beats doing finger sticks.
I know about compression lows, but I just had another No Readings while walking around the house. I’m in warmup now on the new sensor. Tomorrow morning was change day. A sudden drop of 30mg/dl in 13 minutes and No Reading for 35 minutes. I don’t think Dexcom has this extended sensor thing worked out well. This is 5 out of 9 doing the same thing on day 7, 8 or 9. I know if I call Dexcom they will send another sensor. I would rather have an explanation and for them to work for 10 days. When I call them they ask basically the same questions on their troubleshooting guide of the website. I would like to speak with the design team. I know, fat chance.
I am on MDI with a digital pen that microdoses down to 0.1 units of insulin which works perfectly for me. Although the G6 is by no means perfect, it sure beats doing finger sticks.
What pen is that. I have always used syringe and vial doing half units by eye. One thing I noticed about Humalog is it is more viscous than Lantus. Very easy to have a bubble form in the syringe. They are big enough to make a dose off by .25u I guess.
I use Pendiq 2.0. I used to have 1 pen for Humalog and 1 for Lantus. I totally discontinued using Lantus after many years of Lantus giving me nighttime lows.
In a few weeks the G7 will be launched so will see if that one does any better but I would not count on any improvement in performance. The G7 is now in final clinical trials and it appears the the factory which will produce it in Penang, Malaysia is ready to roll. My biggest concern is delays due to lack of shipping containers in Asia which shows no letup in the near future.
I’m positively in love with my T:slim w/ Control-IQ. In 32 years, I’ve never done remotely close to this kind of control. But I think part of my success is knowing Dexcom’s downfalls.
The biggest thing for accuracy is being well-hydrated. I’m one of those ones who can usually run a single sensor for a month with incredible accuracy. I’ve chatted with a few others who get similar life, and the biggest similarities between us are that we all avidly drink our water and generally have excellent TIR. Most people are dehydrated and don’t even know it. You should be drinking half your body weight (lbs) in ounces of water every day. The average man in the US (nearly 200 lbs) should be drinking 100 oz, or three quarters of a gallon, of water every single day. Few people can say that they do. But when if comes to Dexcom, you’ve got to have that extra fluid between your cells to move the sugar around. No fluid, means stagnant readings, and drastic inaccuracies.
Those lost data segments are usually caused by either: dehydration, compression lows (just like with dehydration, there’s no fluid between the cells where the CGM is reading), or less commonly because drastic BG changes have confused the sensor. All but the last one are easily prevented or remedied, putting your automated system back in business quickly.
Unfortunately, the lost data can also mean the sensor has reached the end of its life. In that case, Tandem has an excellent arrangement with Dexcom. They’re pretty happy to send replacement sensors if the one you’re wearing isn’t behaving reliably. They expect the sensors to work, since our lives are literally hanging in the balance.
In the end, this is the only thing that matters. If you can’t trust a system, then it is NOT a good fit for you. The automation can be turned off on all pumps that are capable of it. I loved my T:slim for almost 3 years before Control-IQ was even introduced. I’m an utter technology geek, though, so I’m fanatically excited to be a cyborg!
None of this is true for me. My TIR on several days during the week is 100% and a bad week is when it gets down to 97% with a range of 55-130.
I drink 36 oz of water before I shave every morning and then hydrate all day finishing at least 1, 2 liter bottle so I am hydrating 100 oz+ and only weigh 125 lbs (5Ft 9" tall). I have not had a compression low for years as I place my transmitter in an area of the body that the transmitter never even comes into contact with a firm surface. My BG’s are not erratic, or it would affect my TIR. The sensor is not end of life as it comes back online fine after between 15 minutes and 3 hours for the duration of the 10 days.
I think we can just agree that we are all a little different and what works for one person, does not work for another.
I know that my very low BMI is an issue and Dexcom knows this too as they do have plans in the works to deal with people in my situation but have too many projects in the works to correct this right now. During the coldest Winter months, my weight will go to 130 lbs and between Thanksgiving and Christmas with a lot of family birthdays and other gatherings at that time I can even occasionally go to 135 lbs and these drops mostly if not all go away.
I did say “usually”. I don’t think your particular situation is common, but thank you for the reminder. I have indeed read of others who are also too slender for CGMs. I don’t fall into that category, so can definitely over-look it!
Do you have nowhere at all that you can “pinch an inch”? Arms, top of the bum, love handles?
Nope, any extra fat goes right to my belly, where I can actually pinch an inch, however limited real estate I can use there and as a side sleeper still keep from compression lows.
On the last Dexcom earnings call, Kevin Sayer hinted that he was shooting for an early Q3 surprise and we do know that the G7 is in its final clinical trials now. He has given early expectations before and then delays crept in, so nothing is ever guaranteed until it actually happens.
I’m with @Timothy … I have Dexcom data probably 98% time and, as a result, C-IQ can work properly. For those rare times when feedback is not available, my pump settings are still pretty good.
I don’t need 100% to be happy … heck, even with all this high falutin’ technology, I still forget to bolus for a meal. I’m worried that I’m not sufficiently reliable to be allowed to use Dexcom technology.
