What is the problem with the graph? Is that not what auto is supposed to be doing?
Hang on a bit.
You can do what you want, of course. But rewinding a bit…we were trying Levemir. You were sick for part of it, but it seemed like you were making solid progress.
But then you wanted to go back to auto-mode and give it a fair shake to prove it would either work or not work. And you said you wanted to give it 2 weeks, I think.
There was also the idea of using the pump and going completely in manual-mode.
So you have a few things that you are bouncing around between here. Auto-mode, manual-mode, Levemir, Tresiba and pump, Tresiba and Levemir…
Certainly chart your course and try whatever you want. But I hope you can be comfortable feeling like you have done all you need to with auto-mode and don’t want to pop back over again in a few weeks thinking maybe you missed something.
Every different regimen takes time to get tuned. You have to commit to being able to work through the tuning process for whatever you are doing.
I’m happy to offer advice on whatever you want to try. But I wanted to share my thoughts on it before you jump on something else.
3 Likes
Yes. Yes, yes, yes.
As an observer, it seems to be some new regimen every few days.
If I change one of my basal rates, I give it at least a week and preferably two before I decide whether it’s working. Because there are so many other variables involved.
Try something for two weeks. Give your body time to catch up and adjust. Sounds like a novel concept for you 
, but try to ignore the frustration of imperfect results and ignore the urge to try something completely different. Then you’ll maybe have a better idea of whether a particular regimen is working. And maybe a clearer idea of what might work instead.
6 Likes
That is one of the biggest things IMHO which makes this so very hard.
Sometimes easier said then done. But it is how (IMHO again) it has to be done to avoid chasing your own tail.
Nothing will be perfect. Similarly, you can not compare your results to somebody else’s results. Use somebody with results you really like as a motivator but do not feel like you should be trying to match or catch up.
6 Likes
Dear @Eric, @Thomas, and @Beacher,
I hear you guys, and I understand what I must look like from where you sit. I’ve pumped for 12 years, I went into Auto last year and despised it, got completely submerged in everything 670G, and find myself here, this last month in a full on scatter.
I’ve always been a little crazy, I’ve always been a little over caffeinated, but I’ve never had stuff going on like I do now. I’m at my limit.
I decided on going with shots because I was desperate for a break. I had no idea what to expect, but I really enjoyed it. My mom didn’t. I worry about her nonstop, and she has taken so many hits lately with her health that I decided I couldn’t let her worry about it. Not for the next couple of weeks anyway. I’ve been telling my group for a while now that I would go back into Auto for a trial, and I just thought to get it all out of the way. I didn’t like auto then, I don’t like it 3 days in again, and I don’t imagine I’ll stick around.
I don’t think the Tresiba is going to fix anything. I think it’s going to cause Auto to deliver less. Eric, you’ve mentioned it before, and this will probably be the only time I’ll have the opportunity to try it. I don’t care either way. What I believe about Auto at this point is that no matter what I learn about it or what tricks I use to get it to work, it’ll never be anything more than that… tricks. I hope you’ll understand that while the conversation you guys engage in is in here, the majority of mine is in my room and is about the 670. There were things I needed to confirm. And I have. I’ll probably hang out for a few more days and do a sensor change, since that’s a big problem area for people, and then I’ll get back out. I’ll stay in manual until my mom is on the other side of her stuff, and then I’ll see my endo for some prescriptions so I can try shots again.
I’m scattered. Really scattered. I’m trying to find balance, and it’s not easy.
3 Likes
You should feel no obligation to do anything related for your health for any non-family member.
Doesn’t matter what you did or did not say.
Your health is your own business and that of your family. Do what is right for you.
Family first.
Let the rest of the priorities work themselves out.
6 Likes
It sounds like you are doing an amazing job of triaging right now. I’m serious about that. You’re triaging for everything going on in your family AND you’re trying new things for your own health. That’s something to be proud of.
I hope you don’t feel like anyone here is judging your choices bc I really don’t think anyone is. I also hope you’re not being too hard on yourself bc it sounds like you might be.
We’re all rooting for you. And I think everyone wants to help how they can while also understanding your goals. 
Hang in there, lady.
5 Likes
Thank you, Allison. That means a lot.
2 Likes
This should be fun
1 Like
Wow. I have so much to say I don’t even know where to start. First off, Nicky would feel responsibility to the Facebook group she started, she as helped and inspired so many people who were lost and at the end of their rope. I am one of them. Everyone loves her and really appreciates the work she’s doing trying to figure this beast out. So it’s natural to feel some responsibility, but we the ‘other group’ just appreciate her, nobody wants her to do anything other than Nicky.
Second, If you were a very stable diabetic type or if you stabilize yourself by doing the same thing at the same time every day in the same way, AM (Auto mode) in the 670g Might be be perfect. I am a swinger which means I can go from 30 to 300 in an hour perhaps my basal’s are messed up ha ha… And that Enlite sensor will never catch up with that.
So there are some people who probably already had tight control who just love the 670 G in auto mode. It’s hard to not want that.
I don’t really understand posting in this form but I’m figuring it out. If I hang out here long enough I’m sure to learn quite a bit.
Our failures in life are best lessons.
️:heart:️:dragon_face:
9 Likes
Well @mattie, your post looks great, so if you are having issues understanding the form, you are doing super!
[EDIT] Split this post into a new thread: 670G survey
4 Likes
So I’m late to this conversation, and take this with a grain of salt because I don’t use the 670G but is there any insight you can glean from your pump to figure out how much actual insulin is in your body at any given time in Automode? I ask because some of the problems you describe are also an issue with openAPS sometimes, BUT, I can always fix them because I know the IOB. I know exactly how much insulin Samson has in his system at any given time. I do a lot of manual boluses that are not “prescribed” by whatever the algorithm is doing, and sometimes that means I have to work at cross-purposes to openAPS, but it usually is only partially frustrating because I know all the variables.
Can you find out the reservoir levels at any given time, for instance? If so, you could use that to figure out how much insulin you’ve gotten; just check it at one time and then at another and then subtract time B from time A levels.
From there you can figure out, using historical data (like from your pump break experiment) how much insulin you probably need and give it to yourself manually. I understand the idea that you don’t want to trick the algorithm, but from what I understand the 670G doesn’t use artificial intelligence, it’s not really “learning” from historical patterns in the way most of us understand learning, so it’s NEVER going to know that, for instance, when you have a particular spike pattern at a certain time, that means you just drank coffee and went for a run, and therefore you need more insulin than your last 6 hours would suggest. I could be wrong but that was my understanding from reading about it. So given that, it’s never going to be as good as you are at predicting what will happen in the future based on your behavior. And similarly, it’s probably not going to change your settings too much either. Let’s say it underbolused you and then you gave yourself a manual injection – well, it was targeting 120 after all and that’s where you wound up, so won’t it think everything it has been happily calculating away is correct?
Anyways, feel free to ignore me if this is bad/ignorant advice.
2 Likes
Not bad advice at all! You’re spot on as far as I see it, and your idea to check the reservoir is a great one. It would also take a little organizing, but I’m currently writing down all micros and adding them together. It can’t be a more painstaking effort than that.
I really do appreciate your taking the time to communicate these ideas this morning. I’m shuffling kids off to school now, but I think I’ll make this my morning task.
Thanks, @TiaG!!
I have a question about basal delivery vs regular boluses… if anyone is out there.
First, the pics:
Now, the question:
The pink dots are micro boluses. These are daytime pics from today and maybe yesterday. The picture with the solid line of pink dots was a HUGE increase for me. I’m not sure why, but I had been, for 2 days, receiving anywhere from 100-180% of my regular manual mode basal rate. At the time of the picture, I was probably receiving about 140%.
The other picture, the one with no pink dots, is this afternoon. Most of today has been like this. I think if I added up a couple of hours’ worth, I’d have MAYBE received 10% of my manual mode rate over 3 hours.
I haven’t started the climb yet, though I’m sure it’s coming, and I’m here wondering about the difference in delivery between the two days. Obviously the “basal” amount is different, but so is the amount I have on board from food boluses. I hadn’t done any boluses the day with the high delivery rate, but today I have had as much as 6 units on board.
All of that wasn’t the question. It was the build up. Here it is… Is there a difference between a steady delivery of micro boluses (which I understand is not the SAME as a basal delivery but does REPLACE it) and intermittent large food boluses? In this case, I believe I’m doing insulin for things I wouldn’t need it for in manual, and that insulin is just going towards the missed basal over the last few hours. And maybe this is the way it’s supposed to work, but is there any danger in not having received a regular delivery for many, many hours? If I did NOT eat again tonight, and therefore did not buy myself more insulin, would the accumulation of this period have any great consequence?
I’m not arguing the efficacy of such a delivery. I’m only trying to understand it.
Is there a difference between a steady delivery of micro boluses (which I understand is not the SAME as a basal delivery but does REPLACE it) and intermittent large food boluses? In this case, I believe I’m doing insulin for things I wouldn’t need it for in manual, and that insulin is just going towards the missed basal over the last few hours.
so I think the answer is “it depends.” For instance, Samson gets four or six squirts of insulin every hour to make up his “basal” delivery rate – so if you were giving yourself food or correction boluses around that often and they were roughly the same amount, they would be the same. however, if you’re giving yourself, say 5 units of insulin at 1pm and then another 4 units at 3pm and then 8 units at 7pm, that’s not going to create the same insulin profile – you’l have more upfront and probably less around 2:55pm than you otherwise would if, say, you have a basal of 1 unit per hour.
However, in the one picture you have a whole lot of microboluses – if they’re roughly the same in magnitude to your hourly basal rate I would say they’re mimicking exactly what people get as basal.
So second question. If what you’re doing is bolusing a lot more for food and then you’re 670G is turning off basal, there’s not any danger as you still have insulin in your system – it just might be less even than it would be if you were getting a steady basal rate. However, those phantom rises and falls are, in our experience, worth it for the smoothing out of the highs and lows at other times. But for you that may not be the case…it may be that you’re getting spikier as a result of the on-off of the basal and the irregular insulin delivery.
What you should really do is look at how much insulin is leaving your reservoir. Then look at your normal basal rate, your normal bolus amounts for what you ate, and then look at specific times – over, say, 5 hours, are you getting about as much insulin as you would have if your manual basal rate was going strong and you were just bolusing for food? If so, then I’d say the algorithm is doing an okay job for you. If it’s a lot less or more, then that would make me suspect the algorithm is not doing it’s job well.
1 Like
Funny. I just did a video talking about my last 3 days. I averaged 78% of my manual mode rates the first day, 96% the second, and 39% the third. The hourly fluctuation is mind boggling, too. I have had as high as 184% sandwiched in between a couple of hours of 130-140%. Then today? 2%, 4%…
It’s my activity level that gets it flipped. But that’s annoying. 
I’ve been trying to not pollute FUD with 670G questions… which has been hard… but I have one that I’m hoping someone will explain. I’m on day 16 right now, and I’ve just about had it. It has been very interesting to return to it understanding all I do now, but even knowing it hasn’t been able to prevent me from landing on the same track.
I do have a question though (a million actually) about how this happens. I’ll give you my scenario today and my theory, and then if anyone wants to educate me or debate me, I welcome it.
Not debate… I didn’t mean that.
So this is my graph screen, and the pink dots are all micro boluses. As you see, there are none from 2:30 pm until around 6:15. The small yellow icons at the bottom represent food boluses, and the small blue icon just under the 4 is a correction. That correction was 3.4 units, and the food boluses probably averaged 1.5-2 units each. At points throughout that micro bolus break, I did have as much as 6 or 7 units on board. However, with the exception of the correction, everything else was for carbs that I was ACTUALLY eating (so no “fake carbs”).
I have tried to ask this a few different ways, and I’m still trying to get the wording right, but I feel like these long periods just cause major highs. Sometimes I am even getting insulin, but the delivery is near negligent at 10% or 15% of my regular basal rate. I can see how getting those food boluses on board can help for a bit, but I inevitably rise nonetheless. My thinking is that I’m not getting enough insulin. I can’t go an hour or four without a substantial delivery. I also have it in my mind that those food boluses are a bandaid over top a gaping wound. I have thought about it as an “assignment” of insulin for a very long time, but maybe my thinking here is incorrect. If those food boluses are done for carbs that are being consumed but because I’m doing the boluses, my micro boluses are being suspended, aren’t I doomed to go high? Yes, I have insulin on board, but they have been assigned to cover the carbs. Which means, to me, I’ve got nothing in the background where I should have a steady basal rate. Can insulin cover more than one task? Is this any different than hoping your basal rate would cover your snack?
This isn’t a complaint. I’m just trying to understand. I’m trying to figure out what I would need to fix in order to cover all of my insulin needs… if that’s even proper thinking.
These sensors are terrible, and they are creating even more problems than I experienced the first time around. I believe they’re just making a bad situation worse. However, the trends are still the same, and I’d like to leave this time knowing why it didn’t work.
@Nickyghaleb
What do you see as the difference between a steady delivery of micro boluses and a basal delivery ?
Good question… I guess it depends on the rate and size of the micros. I’ve had bursts where I’ve received nearly uniform micros for an hour or two, but then I lose everything sometimes for hours. If I am receiving a relatively constant delivery of micros, I would consider that comparable to a regular basal rate.
How’d I do? 