#313 Fox in the Loop House - part II
Scott Benner
Kenny Fox talks Loop Management
Kenny and Scott talk about DIY Loop and type 1 diabetes. part 2 of 2
You can always listen to the Juicebox Podcast here but the cool kids use: Apple Podcasts/iOS - Spotify - Amazon Alexa - Google Play/Android - iHeart Radio - Radio Public or their favorite podcast app.
+ Click for EPISODE TRANSCRIPT
DISCLAIMER: This text is the output of AI based transcribing from an audio recording. Although the transcription is largely accurate, in some cases it is incomplete or inaccurate due to inaudible passages or transcription errors and should not be treated as an authoritative record. Nothing that you read here constitutes advice medical or otherwise. Always consult with a healthcare professional before making changes to a healthcare plan.
Scott Benner 0:00
Hello, everybody, welcome to Episode 313 of the Juicebox Podcast. Today's show is part two of a two parter with Kenny Fox. Now Kenny's name might not ring in your head like, oh, Kenny Fox, is that like Michael Jordan for diabetes? Well, no. But what Kenny is, is the data of a little girl who has type one diabetes, who really dug in to the DIY loop. And he understands it in a way that I find it inspirational. Now, Kenny and I are going to walk step by step through every setting in the loop, talk about it and kind of a big picture way. If you're not into a do it yourself algorithm for insulin pumping, I get that you don't need to be I'm not telling you to be. What I am saying is listen to the episode anyway. Because it's just another way of thinking about how insulin works. This episode of The Juicebox Podcast is sponsored by Dexcom. The Contour Next One blood glucose meter, and touched by type one, you can always go to touch by type one.org Contour Next one.com or dexcom.com Ford slash juicebox. To find out more about the advertisers, there'll be a little more about them later in the show. But for now, I think we should get to it. We're going to start right at the top by saying nothing you hear on the Juicebox Podcast should be considered advice, medical or otherwise, please consult a physician before making any changes to your health care plan, or becoming bold with insulin. And today. Also remember that what we're talking about is an algorithm that Kenny and I downloaded from the internet. It doesn't belong to a company, you just completely do it yourself. It has not passed through the FDA. So understand that while we're talking. Now that aside, using the algorithm. And watching the data come back from it and seeing how it reacts will absolutely supercharge your understanding of how insulin is working in your body. I'm telling you to watch a an app, a computer algorithm decide about insulin, it just elevates your understanding. At least it did for me. It's taken me a while to figure looping out and I'm probably not all the way there. But I'm getting closer. And I'm going to keep having these conversations with people who are ahead of me until we're all at the same level together. Luckily, I found Kenny and Kenny understands looping. And that's why this episode is called Fox in the loop house. Part Two. So insulin model, when you have yours, even lower than mine is what's the intention of making the insulin model a shorter time period.
Kenny Fox 3:01
So whenever I try to deal with loop, I'm doing my best to represent reality. So because I know that her incident her bonuses and stuff with good bazel only last four and a half hours. That's why it's four and a half hours. It's not just to get like a performance out of it, it's because I really think that the that's how long the insulin last. The rest of the model is a little interesting. If you see that top option is always that Walsh model, it's always got the line like higher than the rest of them. That's kind of what the pumps use it. It's more like a I don't know, like a straight line steady decay of insulin or insulin is really kind of peaky, right, it's slow to start, it hits hard kind of around an hour ish hour and a half. And it stays strong for an hour or two. And then it kind of fades. And so these other models more more accurately represent like how much insulin is left, because it's kind of more initially and then less later, where that wash model stays really high. And so that's why the that's one reason, even endocrinologist know that insulin lasts longer than three or two or three hours. They like to set people at like a three hour because for most of that curve, a three hour four hour Walsh model will accurately represent how much instance on board until of course, you get to the end when you say hey, it's over after three hours, and there's still two more hours left. And so then people are like, Oh, well, why do I says no more instance on board? Why do they keep dropping? Why is my blood sugar keep dropping? Well, it's because there really is more insulin on board but your pump is lying to you. So you either got to have your pump lying to you the whole time until the very end when zero is zero, or and then you can kind of just round down or you set it shorter and then it's like kind of accurate most of the time and then at the end, you still have insulin on board left for another hour too. So it's a tricky thing.
Scott Benner 4:51
Is this a case of because insulin works in choppy like ways, not smooth ways that you're going To have to lie to yourself or lie to the pump, or have the pump lied to you, however you want to think about it one way or the other. So let's lie on the more
Kenny Fox 5:10
cautious side. Yeah, I think for a lot of NGOs, they use the instant action time of three hours to give people permission to bolus when they're high, like, hey, if it's close to zero, go ahead and correct because after three hours insulins mostly done, rather than tell people, hey, this is how long it's been last. And if you see this number, round down, a better idea would be to use a better tool. But those are limited, right? You can use nightscout has a better, like in insulin action calculator, extra spike, these other open source apps and loop their insulin models they use for how insulin hits indicators is much more accurate than what comes standard in most pumps, where it's just a matter of getting these tools.
Scott Benner 5:52
And what I would do prior to loop is, I just wouldn't pay any attention is on onboard.
Kenny Fox 5:59
Exactly. Yeah, it becomes so inaccurate that you don't trust it anymore. So no one looks at it, which is what we'll talk about, we get to base rates and stuff like that's, that's the the piece people kind of ignore they people look at it and go, Oh, wow, there's a lot of insulin on board when someone's falling, but then they don't really look at it other than that,
Scott Benner 6:15
yeah, I just think of it is, I don't know, like, it's like putting out a brush fire to me, like, I've always got to charge toes ready. But I'm not always squirting it, I turn the water on when I see the flames. And so like it just, you know, I go I put something out, I keep walking around, there's more, there's more carbs, more more insulin. Now stop again, it's just to me. Diabetes is like a common sense. Like you can use insulin in a common sense way. It's when you get out of balance or out of rhythm, that everything gets messed up. And then common sense doesn't seem to apply anymore. Which is why I always tell people like when you get when you start bouncing, the best thing to do is to get low, get down, get steady and start over. Because you're just you're trying to you're trying to grab sunlight otherwise, you know, yeah, like it's just it becomes it just becomes an endeavor. That doesn't make sense. Okay. Kenny, also, by the way, at the beginning of the podcast, we're gonna let people know, this is an algorithm that you downloaded offline. It didn't come through a company it has not by anybody except the people who wrote it, who, by the way, are really brilliant people, but are not the FDA. And that you're not a doctor by any stretch of the imagination, or am I? We have no, not even close. Right. And that none of this is advice. We're just talking through how we do it people should you know, definitely be precaution. You know, take precautions. Keep in mind that when I started the loop with my daughter, I had Jenny to talk to by text, I was able to reach out and speak to Katie, and had I not I don't think I would have made it two days. Honestly
Kenny Fox 7:54
with it. It was it's very frustrating to start to or it could be scary depending on which way you're coming from
Scott Benner 8:01
hundred percent. Okay, so what should we talk about next? Do you think basal rates?
Kenny Fox 8:05
Well, real quick on the model only thing left is that peak, you don't really see it much. But it's mostly like supposed to represent when insulin is at strongest. I use the adult one for my daughter, not the child one. So the child one has like a 65 minute insulin peaking and the adult 175 I feel like a little over an hour's when I see insulin pulling her down the best. So I went ahead and did that. You also see that if you extend the peak out and use the adult one instead of the kid one, you'll often get larger initial bolus recommendations, just because of the way the math works. So I use both a lower insulin action time and the adult peak that's a little higher. a fun thing if you want to test or kind of play with the shorter da instant action time without or peaks. without, you know, really testing it like an open loop or with your PDM or whatever is you can when you do the build. And there's a screenshot in that loop docs on how to change the insulin action time and peak, Katie had a great suggestion, just take one of the models you're not using, for example, we don't use vs. So I took the settings for Fiesta, and I changed it. So I just changed the peak to be what I wanted it to be like matching the adult one for example. And then I change the insulin action time to you know, the five hour one or the four and a half, whatever you're going to test. That way once you build your app, you can just switch between the models, and like use them for what you want to use them for. So if you want to try a five hour interaction time, and then find it's like too much, it's too aggressive or whatever it is, then you can just go into the settings and just tap and use the different model. You don't have to go back and rebuild. Sometimes building causes a lot of stress for people. So if you can do it once then you can just kind of toggle back and forth without having to do a new build. It's helpful. I guarantee you just spoke about most of the people's understanding as well. If they're getting into that Lu customisations. Yeah.
Scott Benner 10:03
Okay. sila rates is good. Yeah, rates is the next place to go. So you just said something, you know what? Maybe it's not maybe those things strategies the next place to go. So that's a cool one. Yeah. And here's why not all of you who even download loop are going to have this setting. So in very basic terms, if someone has written a program, it's an app, right? And this app is an algorithm that takes the data from the Dexcom, and sends it to your insulin pump, and makes these insulin decisions happen around it. There are other people working on additions to the idea. So there's sort of a basic one, is that called the is that always dev? Is that the one that's being constantly being developed? Or am I speaking wrong here?
Kenny Fox 10:51
Yeah, you're right. So the master one is the one that you know, is should be the most stable, it's kind of the one people would mostly download. Typically speaking, Dev is where all the is where some of the playing, there's actually like, kind of sort of a layer beyond that, like a what they call like feature testing, someone might build a little, a little change, they want to try, and they'll build that. And then once they kind of test it, it looks good. And they kind of put it into Dev, and that's where all kind of the fun stuff, the new stuff comes together, and hangs out for a while while people use it and make sure that all those new features don't just work, but they work well together. And then eventually, that gets promoted up to master after it's,
Scott Benner 11:32
everything looks good and proven out. So the very first time someone mentioned to me about an automatic bolusing. So right now, this this, this, this master branch of loop does not does not give you boluses of insulin, it manipulates your basal rates to try to try to get you to where it wants you to be. But I was somewhere a number of months ago, and some guy says to me, yo, have you seen this Auto bolusing? You know, addition to the loop? And I said no. And he told me about it and helped me check it out. And I forget if it wasn't written by a man named Ivan, maybe Ivan. Yeah. And Ivan's was cool. And you could turn it off and turn it on. But the ways to make it work were a little too detail oriented for me. So I was able to make it work pretty well. When there were carbs present in her body, it actually had a setting for like, have this work while carbs are present, or while carbs aren't present. And I couldn't make it work without carbs. And I and I could make it work with carbs. But when I tried to leave it on constantly, overnight, her blood sugar would be like, you know, 80 and all of a sudden it would like be giving her insulin. Oh my god. That's no good. So we got away from that. When an auto brand showed up. Did Pete right it? p pizza and the main steward and developer of lupia. Right. So Pete wrote this one, I gotta tell you, I've been using it for a while now. As a matter of fact, I got a note from Kenny and Kenny, he's like, yo, you seen this developer? This, uh, Audubon springs from Pete and I was like, already got it. And so yeah, you were ahead of it. I was surprised. Sometimes I sometimes I'm out the right out in the forefront. The flag I don't even have a sword. I'm just running ahead with the flag. So I'm the first guy to get shot Kenny.
Kenny Fox 13:24
Someone has to be
Scott Benner 13:25
shoot the guy with the flag first. So. So anyway, we're using that. And I have to say, it's pretty Skippy. So Am I understanding it right that when and insulin is pending, or when the you know, the loop is telling you, we think you're going to need another half of a unit soon. It will give you 40% of that.
Kenny Fox 13:48
gas. Yeah, exactly. So like the by default. So instead of increasing basal rates, it only decrease your basal rate, it will never increase. If you need more insulin, instead of ramping up your temporary basal rate, it will give you a bolus. But then just for safety, there's a a number you can actually change it to in the code, but it just uses 40% of whatever loop thinks you need. And it gives you that as a Bolus in the next five minutes. So I'll give you another 40% of that. So it kind of its slows down. So I mean, in theory, if your insulin need was the same, for let's say half an hour, because that's how long loop kind of can can give you a basal rate for it would give you it end up giving you less insulin over that half an hour because you'd never get to 100%. You know, 40% 40% 40% just gets smaller and smaller and smaller, but never really gets to zero. But it gives you a lot more up front. So Pete did some cool math was basically saying that if you had consistent insulin needs and you're going to run a single basal rate, increased basal rate for half an hour. You're basically getting 17% of what loop thinks you need every five minutes. If you just held steady for that half hour and then but With the auto bullets, you're getting 40% of it kind of right away. So you know, within the first 10 or 15 minutes, you're still way ahead in terms of like timing of the insulin, but after about 20 minutes, you're kind of about the same amount of insulin delivered. Unless you count the fact that we might change the base rate every five minutes, in which case you go back to what I mentioned before, where it starts counting over again to redeliver. So
Scott Benner 15:22
right, so using Arden's current situation, which, by the way, please believe me, no one set up just because it works perfectly during Kenny's conversation in mind right now. Like, so I want you to know, first she comes out of gym, you know, she's lost connection, Toradex calm, while she was disconnected the loop had taken her bazel away completely, so she didn't have any insulin coming in. And she gets back over finds out, hey, my blood sugar is actually you know, 60 I want you to know that we checked that with a finger stick. So she whipped out the Contour Next One meter, she checked her blood sugar, she's definitely 60 she feels fine. You know, I didn't share everything. She said in the tax. She's like, I don't feel dizzy, I feel 100% fine. She was you know, 15 or 20 minutes before that, like in the low 70s. And most of your like, the kid just went the gym class with a 70 blood sugar. But it also is because I knew that for the time prior to that she did not have a glut of even bazel running because this 85 blood sugar that was trying to drift eventually to 60 which it did over a number of hours. Loop has been taking away bazel for a while now. So I was pretty certain that even if she kind of got a little lower, it wasn't going to be a crashing low. It's just going to kind of like float down, which is exactly what happened, which is cool. But I just want to be clear we you know, I my daughter seven units of insulin based on the hope that her CGM pop back on it was right which by the way, it did pop on and it was right she it had her at 64 the the meter header at 60. You know, I'm that makes me comfortable. You might imagine the next reading drifted down to about 60 and then come back up to the meter. Yeah, we're at 60 right now. And but here's the great thing. She's now been eating for quite some time. So just like we talked about earlier with the cereal, like where you know, she looks like she's 85 or 90, but the cereals in there. Clearly, she's going up because she's eating the cereal, but the glucose monitor has not adjusted out of it yet. And we gave her some crushing amount of insulin vino for the cereal. And it still didn't help by the way Arden's coming out of that cereal, she still at one point was 171 error straight up. And the way we stopped that was by opening the loop. And hitting her with a big bolus of like three more units, she eventually had 13 units for that cereal, which, ironically, is how much I would have given her without loop. And I don't know why I didn't just do that it was because it was because I wasn't with her and nobody was around. But we stopped that cereal spike at like 185 200. And that space, it it leveled out. And as soon as it started to come back down again, we close the loop right back up. So the loop could start taking away bazel. Because those three units were too much they were enough to stop the spike. But they weren't going to be needed moving forward. And because ardens bazel rates like 2.5 an hour, I just basically I threw in an hour's worth of bazel to stop the spike, and then took away an hour's worth of bazel. And let that three units act as the bazel.
Kenny Fox 18:25
Does that mean we do something very similar. When I see a spike like that, I just don't bother to open loop because one, I can't do that from the watch. I'm not going to bother my six year old to say Come over here. Let me have your phone. And then I'll do that. But also I know exactly what you said, which is that what I'm giving ultimately is probably too much unless it's unless I miscounted the carbs. So that's one of those situations where it's like, well, if I miscounted the carbs and I'll go in and edit or add a carb entry kind of back in the past to kind of stack on top of the meal. But But even before I can even figure that out, it's just a big bullet. So you just need to stop the arrow, give it a big bolus, let Luke cut bazel I don't care if it cuts bazel or not. And if I think I miscounted the carbs then I go back and I add carbs. If I didn't miss count the carbs and just missed Pre-Bolus or whatever Miss timing, then I just let loose cut the bazel because for the same reason, she's going to go low eventually if it doesn't, so I just let it do it. That may mean you get a couple more ticks up versus open looping. But I just I let it ride. I just give a big enough Bolus to stop it in its tracks and loop cup bazel and hopefully land safer or safe safely or safer than she would have otherwise. I've done
Scott Benner 19:37
that too. I don't it's funny, isn't it like what you just said makes complete sense they but I would have been uncomfortable giving her five units to stop a one at going up. Even Yeah, even though I think you're not wrong. I think it probably would have gone the same way. So I'll try one day like I just like having the bazel back there.
Kenny Fox 19:57
Oh, it definitely helps. Yeah, it makes it makes a difference.
Scott Benner 20:00
So, so now right now at this meal, we've put in seven units for what I'm guessing is like 80 carbs. But at ardens ratio. She has more than that she needs more than that. Now I'm going to tell you right now I've seen it work enough times, going back to Pete's Auto Bolus, those things strategy to tell you that if we don't do anything else, that Auto Bolus is not going to let her go over about 170. That that's what I know is going to happen based on what because I've watched I've let the auto Bolus do its thing a number of times, so I can watch it. And so if there's a if there's a reasonable Pre-Bolus, you're starting with a lower number like this, it's going to stop her up in there, and it'll get her back down again. But the truth is, is that as soon as Arden's blood sugar adjusts, in a way that makes me comfortable that we're seeing some sort of upward mobility, you know, movement in her blood sugar, I'm going to ask her if she finished all of her food. And if she did, then she's going to put in probably three more units on her own put in the rest. Yeah, yeah, it's so it works. And I don't know, the technical side of it is lost to me. I don't know the changes that they've made in the basic Master, you know, version of this since we first started. But when we first started this, I think I've updated the master vert, you know, loop, maybe once or twice in the last nine months, the first two goes of it that I was using, were really, night and day, not nearly as good as what this is.
Kenny Fox 21:32
Yeah, yeah, it was a big, big one was how, how loop assumed that the only pod was delivering those basal rates, how it calculated when to start and stop its delivery. That was the big thing that was off a lot of the Medtronic pumps that had been used before, they instead of they just do it differently, they start, they restart the counter, but then they start a little sooner, without getting the detail to start sooner in like the Omni pod and a couple other of the Medtronic pumps so that and I think they sort of knew about it. But didn't think it was that big of a difference in the amount of insulin delivered and how much difference that would make. But with especially with littler kids like us, it made a huge difference, it would say that she'd have a larger percentage of insulin on board then than she really did, because it thought it delivered insulin and it didn't. So that that was probably the first big change. And the second big change was a the car model how loop expected the impact of carbs the hit, I thought it would be kind of steady like all those 80 carbs, we hit evenly, evenly raise the blood sugar across those two or three hours, instead of like more accurately modeling that most food hits faster sooner, and then it kind of steadily goes down. And so those were the two like significant changes that you probably experienced. And I did too, that sort of made me start to trust the system a little bit more.
Scott Benner 22:55
So let's jump to base. All right, let me ask you, for me, you heard me break Arden's Bolus into two different decision making entries? Yeah, well, actually, it's one it's so she makes an entry. So she said whatever it was, if it was 35 units, two hours, it means it says two or however much it thinks she should have, I always tell her just choose zero and enter it, then get that Yeah, put in the next one that it bundles them all together for the next one, so that you don't have to watch insulin, so that it doesn't deliver the first few units of insulin for the two hour model. And then you have to sit and wait for the to do that. So instead you deliver nothing on the two hour or whatever the first time because you're thinking and the adult delivered on the second go round. But I do that, so that Luke has a better understanding of exactly what you just said, some of these carbs are going to hit faster, some of them are going to stay with her longer. And that used to keep the bazel from going away. And it still would if I didn't have the dosing strategy of auto Bolus on. Because I have auto Bolus on it's now taken her Basal away, but it's going to try to come back with Bolus if it needs it. Whereas if we didn't have auto Bolus on that thing I just did with a two hour and three hour would have tricked the loop into keeping the bazel on.
Kenny Fox 24:15
Yeah, so like that three hour people find what those carbs is that if you put a longer absorption time, the initial impact is less you get less insulin recommended up front, typically you'll get the most insulin recommended up front on a two hour. But on a three hour let's say it's going to last longer, and it's going to the carbs have a chance of outlasting sort of the peak Enos of the insulin and Luke knows that so that it will offer more insulin later to kind of help compensate for that. So sometimes when people find themselves kind of high at the end of a meal, it may not be because there wasn't enough insulin it may just be that you needed more insulin but you needed a little bit later and you seem to totally pay This food actually is lasting longer than I thought. And it will often recommend another Bolus once you modify that absorption time. In the past like, Oh, this one, I thought it was two hours, it's probably more like three, you change it, it'll probably give you a bolus recommendation in most cases. So but doing the like, representing the food properly is isn't is important. And it's kind of a new thing I had to learn with lupus like, I don't know, how long does something take to eat night last in your system? So
Scott Benner 25:26
no, I know. It's interesting. I just, I kind of blank I do two and three hours. And it works every once in a while. If it's something like rice, like with a Chinese food, I might do two, I might do two and four hours. But it's you know, it's not always perfect. Again, I want everyone to keep in mind that Arden's a one c does not come from an 85 line that never moves. Like to be perfectly honest, like, because she was sleeping in. And you know, it all, you know, it all went well for me this weekend. But that's not normally what her it's not normally. Yeah,
Kenny Fox 25:58
it looks like for eight hours, you know, and I think that's part of the secret sauce or magic sauce of like a control IQ or some of these other algorithms is they you just enter carbs. And you don't talk about absorption time even like open APS and other systems, they have other algorithms, they tend to handle the dynamic carb things looks like a little bit better. You don't have to worry about how long the food is necessarily. I think that's that's pretty magical. But loop is, lets you kind of see all the pieces like it helps you understand, like, biologically what's happening.
Scott Benner 26:28
I think that that's really why I'm telling people about I always talk about glycemic index and glycemic load with people like you really need to understand the impact of the food. You know, it's how, how hard is it hitting? How long is it hitting? How long could it go? You know what I mean? Like, is this a? Is this a food that can only make it two rounds? Or is it going to go the full? You know, is it gonna go the full way? Is it gonna crush you? Is it gonna come out like Conor McGregor? And just like jump through the air and, you know, jam its face into yours? Or is it gonna go a little slower and you know, start more of a seductive dance with you before it starts the pummel you.
Kenny Fox 27:02
Yeah, like juice or candy or glucose tabs like I enter those 30 minute absorptions. I mean, they're probably more like an hour, but I do it as 30. So that loop doesn't try to overcorrect those. But I mean, generally speaking, like juices, kind of a hit quick and go away fast. You don't want to put a play introduces a two hour and she's just having it just for fun. Or having grapes by themselves. And that's it. Yeah, they don't last two plus hours and loop will end up making her golo kind of later that when you know because it just thinks the carbs are gonna last longer, and they don't.
Scott Benner 27:34
So let's talk about the basal rates then. And I guess we have to talk about them in two different ways. Like, do if you're using an auto Bolus pizza? And if you're not, so if all right, I don't know, though, it's funny. I have no
Kenny Fox 27:46
basal rates to just bazel Yes,
Scott Benner 27:48
I haven't changed ardens bazel rates, since we went to the auto mode, the only thing to understand is that is is how it makes up for stuff, it doesn't make up for bazel anymore, with anymore.
Kenny Fox 28:00
So I think this is where this is where I think this is where I start with most people that I end up helping out is uh, basically you know, is like switch your body needs some and you and Jenny have talked about that multiple times. So it's like if you didn't eat, or run around or whatever, and just kind of fasted all day and watch TV or something, you should be able to stay flat. Same with overnight. And so that's kind of the goal with Basal. And so but what I've found out get into like how you track it here in a second. But what I found is most people have like one bazel rate, maybe a couple but if you start with one, you end up able to see when those other changes show up. And I came to this just because on shots, Tessa was level that the basal rate and the insulin on board calculation worked all day, every day at any time. Day or night, it would always end up zero ended up being zero when she had no insulin on board. It was everything was done. So I just came from the mindset that she only had one basal rate, but I got the pump and start playing with it. And it just would mess up the math for tracking how long the insulin was working. So I just kind of stuck with one. But I found that with other people, it works pretty well to you just pick one. And then you can kind of see if you need another one. But you start with one I guess I just heard the other day that the sugar surfing guy, Dr. Ponder, I think sort of thinks the same thing. And what I think actually changes throughout the day, which we'll get to in a minute is sensitivity not bazel but when in a world of dumb pumps, all you have for automation is basal rate changes. So I think people are in one sense programmed to think oh, it's probably bazel but really what I think doctors and people are compensating for with base rates is often sensitivity changes not bazel but yeah, so Basal is is what is supposed to be but loop sees Basal as neutral as free insulin as zero doesn't track it. It trust that you said hey, I need this much insulin. Kind of as a baseline for my body. And so it doesn't really keep track of that. So if you have bazel too high, you may end up with the insulin on board number may show that you have zero or maybe even a negative number. And really, you're falling you actually have more insulin in your body than loop is aware of. But because I trusted you, I said, Hey, you said you need to this much insulin all day. So like, that's, that's what I'm giving you. So when you do an override, and you increase those bezels, it also can kind of complicate that calculation. So yeah, I think that's kind of the main thing is, is Basal is free, so you don't really track it. So what I do to test bazel, with closely if I find the search with like the tandem system as well, is you can test bazel without testing bazel in a closed loop. And it's pretty awesome, because you can look at that insulin board calculation, you know, assuming that interaction time is reasonable, anything six is a fine place to start six hours. But if you look at the insulin on board, overnight, you should get to kind of where you want to be in that correction range we talked about, and iob needs to be pretty close to zero and flat. If I obese, not zero, meaning like just bazel is all you have running at night, and you should be flat. If it's not zero, then something's wrong with your bazel. If you're constantly have positive insulin on board, like a bigger number more than zero, then your basal is probably too light, because it loop is constantly having to add more insulin to kind of push you down into your range. And if you're especially if you're above the range, you want to be insulin on board, that's a positive number, especially at night should always mean you should be falling. But if you're not, then something's not Luke doesn't know what it should know. And then the weird one that really throws people off is the negative insulin on board when you see a negative number in there.
Negative is is a deficiency of insulin, you and Jenny have talked about that a little bit before where if you turn your let's say your level when you wake up at 80, but someone wants to be at 100. When they Bolus they could turn their pump off, they could do a zero beta rate for a little while and you'll drift up. That just means your body has less insulin than it really needs just to hold you level. And that's on purpose. So negative insulin isn't a bad thing. But when you see negative insulin on board, you should always see you know, giving some room for sensor lag, you should see either right away or pretty close. Blood Sugar readings should start going up when you see a negative insulin onboard situation. If it's not going up when it's negative, then you're not representing the insulin properly in the body and loops gonna loop thinks you need more insulin you don't. So what happens oftentimes is people are falling, the blood sugar is falling and there's negative insulin on board means their bases are too strong. So as soon as you start curving up, let's say you treat a low loops like oh, yeah, I expected you to come up not knowing that you actually gave carbs. And so it's going to try and fill in that negative because what should work in this situation I talked about where you turn your pump off before breakfast, you actually can't just turn your basal back on, otherwise, you'll still keep drifting up because your body is missing the insulin, it needs to kind of maintain that balance. So loop tracks that negative amount how much you're missing. And it's going to fill it in for you to try to level you out. And then also correct for any, you know, upward momentum. So what ends up happening is you know, getting slammed back down, because you have too high of a bazel you have negative insulin on board and you're falling, and then you correct but don't tell loop about it. And then it's Oh, good, you're coming back out. Let me give you more to level you out and it's pushing you back down again. So you end up in this cycle overnight, we're constantly trying to bring someone up and lips pushing them back down because he thinks you're missing insulin, but you're really not. And that's what that negative insulin on board will tell you. And so you have to adjust your bazel. So that insulin on board is 00 means level, negative means up and positive means down, obviously, without food. And that will tell you so you can kind of scan your day and look for points in time where loops thinks you're kind of around zero and see what the behavior is of your blood sugar. And then you can know if your basal is too high or too low, and starting with a single basal rate will help with that because the insulin runs for four or five, six hours. So feel really high basal rate to catch a nighttime rise like most kids do. And then you cut it back later the insulin that you gave is still running you know until four or five in the morning and it's you're still gonna end up with this like negative iob and falling kind of situation. And it's not the not because your your little basal rates during those eight hours are are too heavy. It's because the heavier when you ran from like 10 to midnight, was probably a little bit too strong and it's just causing an impact later on the night. That was a lot but hopefully that makes sense.
Scott Benner 34:48
No, it doesn't. And I think it's important for people to understand the idea of like sensor like the CGM is reporting behind time a little bit.
Kenny Fox 34:55
A little Yeah.
Scott Benner 34:56
So by way of an example because you spoke there for A few minutes when you began speaking, Arden's Dexcom was telling me her blood sugar was 58. But it's not because she's had food in her for 35 minutes. Right. And so I'm not panicking. It's, even if it's, it's not 58. But let me jump to the end of the story, but it was stable at you know what I mean? So I know this food is now going to do what I expected to do, right? Like trust what you know is going to happen, it's going to happen as food went indoor, it's going to start impacting her as you were speaking, the Dexcom flipped over to 66. So obviously, her blood sugar didn't magically go from 58 to 66. In five minutes, it has been trending up. Like you said, it's now gotten two reports in a row to its to the text comms algorithms now gotten two reports in a row that it believes. So now it's going to start reporting it right. Exactly. Yeah. And so as soon as that happened, it took a little bit of more time for nightscout to know it. So I can kind of see it on a on a different screen, someone watching everything while we're talking for this explanation. As soon as nightscout knew her blood sugar was 66. It put her bazel right back on again. Like immediately, and now I shouldn't be right. And now in truth, this is the moment we should be bolusing that other insulin right now, whatever more insulin I believe she needs for her food. This is the time to put it in probably sooner even. But now for I can be certain with the data. I have backed it. Now's the time. But we're gonna let it go for a little bit just to see the loop does to let let it see what loop does. Now if she jumps from 66 to 90 next time or something like that. I'm bolusing right away. Oh, yeah, maybe even a little bit extra, right. But if she just drifts a little bit here, I'm going to be interested to see because the auto bolus is going to kick in because we only use seven units for something we told her that we told the loop and the loop believes needs 10 units. What's your suspend set up? 6060
Kenny Fox 36:55
Okay, yeah, so it will it could even start giving. That's why it turned the bazel back on. Okay.
Scott Benner 36:59
Right. Yeah, I used to have it at 55. But I she was getting low too much when it was 55. Where's yours? 70.
Kenny Fox 37:05
Okay, yeah, 70. Because again, bazel and curvaceous are locked into the base of the big one. And you got to make sure you you believe that loop will drop you the other big one will be sensitivity we get to but yeah, if you're, if you're not getting down where you want it to be like a lot of people will drop their suspend lower. And I think 60 or 65 is, is fine for the most part. But I find if I'm trying to push that kind of ad overnight, and lose my overcorrect a little and then I get woken up at my 70 alarm. So I just put the suspend a little higher so that mostly so I don't get woken up as often.
Scott Benner 37:38
I believe that I listen, I believe in our basal rates really well, like Arden got up this morning. 630 she didn't eat anything until, you know, just now. Right? And so and you know, she got down to 60 with Jim. So I believe in our basal rates. Perfect. Yeah. And overnight. She's, it's gorgeous overnight.
Kenny Fox 37:59
So I think what's nice about loop and that neutrally makes us all nicer people and talking about having tools is you have an insulin board calculator, right there. So you can see in nightscout, especially, but you can see on your phone too. It's called active insulin in the loop app, as you can objectively measure your basal rates, which is nice. It's not dislike, well, I think it's fine. It's you can at any point in the day, if you see zero and she's not level, then you can kind of question without the bezels right. And I think the other fun thing that comes into play is that idea of school for a lot of kids maybe makes them go up oftentimes, like stress or something maybe work for some people. Once you I'd like to try to tell people get your basal rate, like create a baseline where like to say, like weekends and not school will not work. So you know what normal life is so that you don't have to worry about crushing someone you know, on a weekday, if you can avoid it. And then you play with overrides the override feature increases your basal rate or decreases it however you set it. And use overrides for school and work that way. You can if you feel like maybe you're not as stressed at school or at work one day, and you'll end up meeting it you can just simply cancel the override and kind of go about your day and be fine. But if you need it, then it's there to use it and you don't have to keep fiddling with your settings every Sunday night or Friday night. Things like that.
Scott Benner 39:19
Okay, let's jump ahead here. So to insulin sensitivity now, yeah, it's funny. insulin sensitivity is wildly different for so many, you know, for everybody. Right? Ardennes during the day is 59. And overnight, it's more like 64. But I've spoken to people who have their sensitivity, like you know, it's 120 and so can you. I'll tell you how I think of insulin sensitivity and then please you tell me how you think of it and you know, etc. in my mind's eye, it's just sort of the amount of insulin it takes to impact me. You know, mean like just like this, what what's gonna knock me over? Like, you know, you can give me 59 and 59 will do what it's supposed to do. But if it was at 6565, wouldn't do it 65 would be like getting shoved by a six year old. But 50 nines, you know, like me pushing against somebody who's my, my weight, like, it's, it's the, and I know this is it's not it's not clear, because I'm never clear on insulin sensitivity, like, you know, I'm sure there's a really technical way to say it, I'm sure you're gonna say it in a second. But I find a lot of people get confused by it, and no more so than the idea that a lower number is more powerful.
Kenny Fox 40:39
So yeah, that's weird,
Scott Benner 40:41
right? So a lower No, it's a very basic mathematical idea. But still a lower number is more powerful. So my daughter, five 737 pounds, insulin sensitivity 59, during the day, your daughter, you know, much less. You know, wait, I would imagine
Kenny Fox 40:58
what's hers that? So she's her sensitivity during the day is about 200. And I would probably say realistically, it's probably about 225 250. So it's probably more sensitive. But I've looped dialed it down just a hair, because to make it behave a little more the way I'd like I use that as kind of a, an aggressiveness number to some degree to make loopback a little bit faster than maybe it would normally without causing so much of a low. Yeah, so sensitivity is yes, how much insulin it takes to move you a certain amount of points, your your blood sugars or how much insulin it takes to move you. So yeah, I think it's, it's, that's a good way to look at it.
Scott Benner 41:40
So how so I guess the next question is, and I want to talk about overrides more specifically, but the next question is this. Once I found we had Arden's settings close in loop, then I began to adjust them sort of like an equalizer on a 1991 rack system. Okay. Like I'm like, I don't know what treble point oh, nine is, but what happens if I push it up a little bit? Yeah. Right. Or what happens if I push this down a little bit like that? I, once I was close, then I could start fine tuning without knowledge, then I could start going correction range. Let me try 87 to 85, whoo, that was better. You know, you know, bazeley I wonder what 2.3 does versus 2.2. That kind of stuff. But starting out. You cannot adjust loop to ear? Did you know what I mean? Like it's just you have to start somewhere near? Near good. Somewhere near good. Right. And so my question is, do you know if I came to you today, and I was just like, guy on the street and I say, Kenny, hey, I really want this to work for my daughter. Do you know how to help somebody set this up? Like do you know how to go from scratch? Quick kidding. Add today, the dexcom g six continuous glucose monitor. You want to check it out. It's at dexcom.com forward slash juicebox. type that into your browser right now. If you're you know not good at typing, click on the link in your show notes. It's right there in your podcast player. There's notes in your podcast player, just find them and click last thing you could do is go to Juicebox podcast.com and click from there. All of these options are viable ways to get to the sponsors. While you're there, check out the Contour Next One blood glucose meter, go to Contour Next one.com there's a little button at the top, you can find out if you're eligible for an absolutely free meter. This meter is by far the most accurate one that my daughter has ever used in her entire time with Type One Diabetes. Lastly, if you'd like to see some lovely people doing wonderful things for people living with Type One Diabetes, check out touched by type one.org I'm doing you a favor. I'm shortening up all the ads today. You can do me a favor and click on the links dexcom.com forward slash juicebox Contour Next one.com touched by type one.org that's all I'm asking today. No big sell. Just go check them out. Look I'm done before the music it's like you owe me almost tears there's so much time left it's kind of weird, right? touched by type one.org helps people living with Type One Diabetes amazing organization. Contour Next One best blood glucose meter I've ever seen. Dexcom g six continuous glucose monitor game changing technology for Type One Diabetes there. I said I was going to tell you more about it. Weird. Do you know how to help somebody set this up? Like do you know how to go? Yeah Rach
Kenny Fox 44:44
for Yeah, so it took a while a sensitivity was a weird one to wrap my head around to so the my process is look at basal rate, subtract iob. I go back to someone's graph if they've been running it for a little while. Helps I find closed loop is helpful, more helpful than an open Id be a little harder to see it on normal graph, but you could still probably get close. But I like to find how the neg makes sure that negative means up and positive means down and that they're kind of at their range, where they're supposed to be especially overnight. And then we just try to find what I call like that one magic bazel rate. And we pick one that works mostly overnight is easiest way to observe it, run that one all day. And then make adjustments from there we carb ratios and sensitivity sensitivity. I mean, it's so hard to like, really, I'm not going to stop my six year old and doing a sensitivity test that's even longer than a bazel test. So I once we get overnight solid iob, zero means flat, everything's dialed in, I just tell people, hey, turn your sensitivity more aggressive to a stronger a lower number, until you start to see that line that's fairly flat for you start to wiggle start to go up and down and go above your the range you set. And below that range, you set it to make sure it also goes below but doesn't go below, then it's um, you might still not have a strong enough bazel. But basically turn the sensitivity too strong to where you can see that it's obviously too strong and then kind of back it off a little bit. And that's kind of what I use for daytime ISF because I find that people are more sensitive during the day than they are at night. So you find what works overnight, make it too strong and back it off a notch. And the reason why people are more sensitive during the day I think is because growth hormones exists kind of overnight for everybody, me included, were growth hormones, I was reading some studies on this trying to figure out why this rise at night happens for my daughter that just kills me. So what I said earlier is that no matter how much I bolus her during those rises is IRB of zero was still zero. So the interaction time worked well. That seemed to indicate to me that she actually only had one bazel rate or not, it's crazy, intense rise period. So what I found is growth hormones don't make you go up necessarily, but they do make you more resistant sensitivity. But that means something else must also be pushing you up, which I've kind of found a correlation between how close dinner is to sleep or how heavy dinner is to sleep. And so it's probably just this weird digestion process you have when you're sleeping, that stretches out how long the carbs last, and the carbs kind of have a more of an impact because all these growth hormones released during your REM sleep cycles during the early hours of sleep. And so you have more growth hormone in your system at the beginning. So there's just more resistance there. So the only way to really model that and loop is to make the sensitivity, lower a lower number in those first hours of sleep. And then I sort of like ladder it softer and softer or a higher number as the night goes on to sort of mimic the idea that there'll be less growth hormone in the body. And so I might start with like a 200 during the day, and go all the way down to like an 80, which is less than half of what her sensitivity is, you know what, like nine or 10 o'clock when I know she's gonna be asleep, and then ratchet it back up. So you know, 115 131 6180 and then when she wakes up, it's I usually have around 200 or so. So and I found that to be helpful and to help Luke kind of semi automate and for some people automate those nighttime rises, but it's really not a bazel change as much as we're programmed to think, Oh, that's a rise, I need to get more bazel if you use increase bazel there instead of sensitivity, you may find that negative iob and falling kind of at three or four or 5am instead of just making a loop more aggressive because that sensitivity is really only comes into play when you're out of range. Do you that's kind of the cool part about sensitivity is if you're in your range at you know nine o'clock at night, and she's asleep, and great loop doesn't do anything. So if I made the ISF too strong, it's not a big deal. She's in range. But as soon as she's out of range and loop kicks in, you know, puts the pedal to the metal a little bit more. It could be an x faster.
Scott Benner 49:05
Yeah, it could it could be but oftentimes it's not there in those first hours. So it's, it's, as you mentioned, like how much insulin has taken knock you over. So it's but I think it varies throughout the night more than the day. But that's kind of how I approach sensitivity. It's interesting that just because your child is younger and growing, you're having the exact opposite experience night today than we are. Get like Yeah, a little bit. Yeah, when Arden's asleep, she just doesn't need. She needs less sensitivity, not by a lot, by the way, like listen to what you're doing, you're going from like, you know, in the hundreds to double digits and you're bounced around, we're really only moving between, like, you know, 59 and 65. Like it's not it's a pretty
Kenny Fox 49:48
tight tolerance. But how many basal rates Do you have during the day?
Scott Benner 49:51
Just one,
Kenny Fox 49:52
just one. Okay. Yeah, yeah, one during the day one at night. Yeah, I've seen that. If someone's not really going through a growth period I've seen unit with me too. So I wore a sensor for a few weeks. And I sort of like helped prove this theory out by, you know, all for the case of science, having a lot of ice cream before I went to bed, and I could see my blood sugar rise and sort of stay up and kind of in waves like kind of matching my sleep cycles, kind of stay up a little bit longer and take longer to come back down then if I didn't eat something right before I went to sleep, so I was able to sort of mimic the same process as my daughter, but unless eXtreme Scale one because I'm not type one and two. I'm not growing like you said the volume of growth hormone in my body is probably much less than someone who's actually growing. So the sensitivity will be different.
Scott Benner 50:39
We're growing just not in the same way.
Kenny Fox 50:40
So yeah, exactly. When you're having ice cream at you know, eating like a whole bunch of ice cream at 10 o'clock at night. That's Yeah, we're definitely growing that stuff not to be questioned, I guess
Scott Benner 50:49
I am pretty certain that I'll be wearing a Dexcom Pro in a little while.
Kenny Fox 50:55
So that I well, then you should for science, eat something really yummy before bed and see how that happens. I will be
Scott Benner 51:00
Yeah, I'm gonna be wearing the same thing. Like I just I was talking to Rick Doubleday the other day. And I said to him, when we got done, I was like, I'd love to wear a sensor for a while if I could, you know, to really understand, you know, better what I'm looking at. And
Kenny Fox 51:14
so I think Neil is fun is did this and you should eat something similar to what Arden does when she eats it. And you'll see the absorption time play out in both of you. Yeah, it'll look a little different. But you'll see the food stop at kind of the same time.
Scott Benner 51:28
That's exactly the stuff I'm excited to look at.
Kenny Fox 51:31
It's amazing. Yeah, we ate uh, you know, again, for science ate a breakfast sandwich from McDonald's. And it lasted a long time on those chicken ones. And, yeah, you could see kind of the initial carbs. And you can kind of see the sustained fat and protein in both of us. And actually, in one of those cases where we tried it, her blood sugar ended up way better than mine, on average, so. But it's fun. I do think
Scott Benner 51:53
there's many times where I'm doing a better job for art, and then my body's doing for me. Yeah, so interesting enough for art. And by the way, is her sensor went to 75. And then the next adjustment was only to 79. But it did just Bolus 1.15 units a little while ago, then the 79 jumped to 94, diagonal up
Kenny Fox 52:18
high. So you're above suspend, and it's like, Hey, I know you need more for the carb ratio. So it gave it to you. And it popped
Scott Benner 52:23
on like, so I'm gonna see if I can raise her with a text. If I can, I'm not gonna bother. But if she answers this text,
Kenny Fox 52:31
I mean, just take the recommended bullet, because all you're gonna ask her to do
Scott Benner 52:34
it, ask her if there's any insulin that's being recommended right now.
Kenny Fox 52:37
And you can actually you can try it hover over the loot pill with your mouse. And you can see at the very end of that little hover box that pops up, it'll tell you if there's a recommended bolus.
Scott Benner 52:48
Well, so the little loop thing I have right now says, See, I don't even understand this, the loop has put your
Kenny Fox 52:56
mouse over that box and you should see another box pop up.
Scott Benner 52:59
I only have it on my phone,
Kenny Fox 53:00
I do with my Okay, then tap this tap on the loop pill and you shouldn't need so you can refresh that page if it doesn't pop up right away. But p tap on the thing that has the loop put the number in the squiggly line, you shouldn't get a little hover box to pop up.
I don't know I said you need somebody to refresh it.
Scott Benner 53:18
It's my fingers too fat.
Kenny Fox 53:20
That can happen to a lot of pills. They're crammed together. little boxes just says loop device loop. You should see something like on the hover it'll tell you how long ago the Temp Basal that it's currently running insulin on board carbs on board. Predicted minimum and maximum.
Scott Benner 53:36
Let me go over to my iPad. Let me see if I can make that work. Yeah. Yes.
Kenny Fox 53:51
So bring it up on your computer and you can look at it and it's easier
Scott Benner 53:54
to figure out how to do that. I got I don't Yeah, I don't even remember how to bring it up on my
Kenny Fox 53:58
computer. I'm remember her sight. Are you kidding me?
Scott Benner 54:01
Wait, isn't it at the play here it is.
Kenny Fox 54:03
Roku app.com
Scott Benner 54:14
typing, which is always the best thing to do on a podcast. Yeah, people love it. And really, they're huge fans. Oh, here she is. She said what? Okay, so I'm going to ask her is the loop recommending any 1.8 do it. There you go. There we go.
Kenny Fox 54:48
Now it takes all of it instead of auto Bolus would just give 40% of that right. And then again,
Scott Benner 54:53
which it would likely do again the next time the CGM turned over correct. Yep. So Yeah, there you go. So when, listen, I you know, we're not going to be recording when this is all said and done. I'll remember what happened. I'll tuck in at the end. But look at what happened like Arden's a Urban's blood sugar has been right around at all morning five or six hours that she has been awake. She went to lunch, or she went to the gym, her blood sugar went to 60 she just ate a bagel grapes, puzzles, a chocolate chip cookie, popcorn. And a What did I say? tangerine? Something like
Unknown Speaker 55:32
that. Clementine. Yeah, time. Right.
Scott Benner 55:34
And she ate that stuff. It's 1255. Right now here. And that conversation began. Back here.
Unknown Speaker 55:46
Noon.
Scott Benner 55:48
Yeah, two minutes after 12. So it's, it's, it's, it's almost an hour later. And we have gracefully gone from 60 to 95. And even if she makes it to 150, which I don't think she's going to normally she wouldn't be coming out of a low, we would have Pre-Bolus sooner, and this wouldn't be happening like this even. But this is astonishing for the situation. You know, but if you don't understand how this thing works, none of that's gonna matter. Like it? Yeah, for sure. It's damn near impossible. And so, but I also think that for those of you who aren't considering any kind of closed loop system, whether it be loop or tandem or horizon, when on the PI puts it out, or you know, whatever else. Understanding what Kenny and I are talking about here today will help you make better informed decisions even without a an algorithm, I think, because just watching and you heard Kenny allude to it earlier, just staring for a little bit and watching what happens is such a teaching thing. And when I watch Lupe, take bazel away, give bazel back make a small bump with now with the the automatic bolus or before the automatic bolus when it would kind of ratchet up the bazel. It was fascinating to watch loop take bazel from like 2.5 and make it seven but only for like eight minutes and then bring it back again. And like I was like, Oh my gosh, this is really I'm learning a lot from watching that, you know,
Kenny Fox 57:11
yeah. And like I said, we had slightly better numbers in terms of standard deviation and a one C and time and range. When we were on shots, because of the same principles that once I figured out how to apply it to loop. Yeah, it'll just make your life easier. Being able to track insulin on board is really kind of magical in terms of figuring out your basal rate and just knowing when, when the drop stops, you just got to know and how much to correct for and, and all that it's just it's really empowering or most people are just kind of, again, being dynamic about how you talk about in with juicebox. But like at the same time, if you kind of know when it's going to stop, you can be aggressive and be more aggressive because you know how much insulin is working? And if you overdid it, you know how much to correct for a lot more precision in that approach.
Scott Benner 57:56
I think that loop is going to represent, you know, I shouldn't even say look, I think that algorithms in general, but for everybody, but for us specifically, I think it's going to represent a mid fives a one C. Just base I think so what I've seen and what I know.
Kenny Fox 58:13
Yeah, as I tell people like if they're in the sevens now like if you can just get the settings right and moderately Pre-Bolus then, and don't learn not to overcorrect, you'll get sixes pretty easily it shouldn't be that that complicated. And that's what I'm seeing with some of these people, once we figure out especially like, people that have their kids and they get really stressed at school and just need a lot more insulin these overrides work super well for that. And then, and then they ended up having, you know, like I Oh, there's one person I'm working with now that was in kind of a seven range and is now looking trending toward kind of a 6.2. You know, and and they just have even before that before their stats changed just by making some of the changes I talked about. They just have more, it's more predictable. And now it's not as crazy, you know, at least when they're high kind of know why it happened. So. So that's pretty great.
Scott Benner 58:59
It's amazing. Last thing I want to bother you about is overrides. And I don't know that I think about them correctly. So let me just tell you how I think about them. And you correct me if I'm wrong? Sure. I, I know I don't understand exactly what they do. But when I said an override for art and whether it's a decrease like you know, instead of it 100% of, you know settings, I go to 80% of settings or 50%. That's a decrease in insulin power. Or vice versa. If you go to 120 It's a 20% increase, like that kind of thing. Is it just an across the board increase? Is it literally like the correction range stays the same? The suspend threshold stays the same. But this is an increase of bazel rate. insulin sensitivity is the heart rate show carbery show is that in
Kenny Fox 59:49
Yeah, yeah. And the kind of the tricky thing about using them is that once you turn one on any like entries like carbon trees in that time have that Change applied to them. So if you used too strong of one, and entered carbs or a low one, some people get a little too extreme and they go, Oh, they're falling. So I turn an override on, which is not the time each and an override on but they'll do like a 20% or something really small and then enter like lunch. Well, then it thinks that lunch needs 20% of the insulin, your carb ratio, like 80%, less than what you'd need. And so there's no way to really go back and fix that. So you, you'd have to basically know that that happened and add 80% more carbs to that carb entry to get something equivalent. That's a tricky part. But yeah, an override changes everything, which is good and bad. I think there's a lot of cases where you only want to change one or two things, but because bazel and insulin on board, that bazel calculation is so important. I try to tell people that the override is should only be used in cases when bazel actually changed. Otherwise, you end up throwing off the math if you use like if you're stuck high and you're frustrated, and you use a 200% override 100% more than normal, you're doubling everything. Later on, you're probably gonna end up loops gonna think you're gonna land and you're not because you actually have more insulin in your body than you told it because all of a sudden, you told that your basal rate doubled. Yeah. And so yeah, you end up with a problem where lube would have landed you if the math had been right, but you won't, because you lied to it for a little while. So
Scott Benner 1:01:24
the only time the only time I really use it is sleeping it. So like I said, Arden's like 1.4 overnight cheese 2.5 during the day, that 2.5 kicks in at 7am. So usually she's drifting, I have Arden pretty consistently drifting to 70, when she wakes up in the morning for school at seven. So if I set an override at 6am, I set it about 50. So about 50% of power, this, that's enough to get in front of the drift, so she doesn't drift to 70. And so that when 7am comes, and the insulin sensitivity goes from 64 to 59. And more importantly, I think in that situation, that bazel bazel jump up, it keeps the bazel back at overnight,
Kenny Fox 1:02:06
and yeah, 50 percents about what your nighttime is compared to your daytime, right.
Scott Benner 1:02:09
And that's and that's how I do it. Now, the longer she sleeps, the less she needs. So if she's just going to sleep till nine o'clock that works, but if at nine o'clock, she's still sleeping, it maybe has to go to 40 or 30. Like you really, because at that point really consider what's happening. She has not had food, and she's not having any like body impacts on her blood sugar. Now for sometimes eight 910 a dozen hours like in a really sleeping in situation. You keep taking it away and taking away. The key is that when you wake up, it's got to go right back on. And when you Bolus, like you said before, it's not just for the carbs. It's not it's and it's not unique to Pre-Bolus still, you're not just Pre-Bolus for the carbs. You are you're also have to replace all that deficit that comes behind because basically, it's a paper tiger at that point, right? Yeah. Like any food you take in is just going to overwhelm you because there's just no insulin happening in your body. That was just enough insulin to basically keep, you know, like, like a, you know, a corpse from blood sugar going low. It's just not a living person. Right then nothing's happening inside of her in this scenario.
Kenny Fox 1:03:18
Is Arden have a drop in the morning before she wakes up? And then like a rise after she wakes up? Typically? No, not at all.
Scott Benner 1:03:25
Okay, that's good. You see that sometimes my settings take care of all of that. Okay, yeah, um, it used to be before loop. She'd wake up in the morning at seven. When her Basal went up with Bolus in the morning with Bolus, the rise.
Kenny Fox 1:03:40
Like she does have that rise. Yeah, she
Scott Benner 1:03:42
has it but we don't see it any longer. It doesn't actually happen. Because we're basically constantly like I said, you know, we've been talking for a while, like I said a long time ago, because I'm thinking about now as before, not now for later. And yeah. So I'm, I'm a head of that I Pre-Bolus. spikes I Pre-Bolus rises, if you've heard me talk about, you know, like, extended bonuses. People talk about extended Bolus is one way the way I talk about them is Pre-Bolus in the food, and then Pre-Bolus thing, the spike. Yeah, exactly. Yeah, just always. It's a time travel movie, you know, what's going to happen in the future. You're just you're just always bettan you know, on the right team to win because you've got the almanac, and you know, who's gonna win? So
Kenny Fox 1:04:25
yeah, so when I use overrides, it's really like bazel changes. So what what's kind of fun with overrides if you have your basals dialed in, like I said, using the IP math, and typically people have the same bazel overnight as during the day but not always. If it's good, then what should happen with overrides and a lot of people don't experience when their settings especially bazel is not right. Is this doesn't work but if you can look overnight, and my daughter typically runs around 80 to 90 overnight, and if for some reason she's not if she's hovering around 100 or Hundred and 15 with some positive insulin on board, that means that her, her insulin needs have shifted maybe just for the day maybe because she's sick, I don't know. And I'll just run an override of about 1020 30% increase. So 120 or something like that. And then I'll run it. And, and I'll leave it on for breakfast and breakfast is sort of like the maker break like was this for reals or is just temporary. And if she doesn't go low from an increased override at breakfast, then I run that override all day long until I see her until I see that negative iob and blood sugar falling sort of situation again, where things sort of settled back at over. So I just roll over out of bed, look at nightscout see where she was at. If I get woken up with a 120 or 130 alarm overnight, and Luke can't keep her under that number, then I know her insulin needs have gone up a significant amount like she's probably sick. So she probably needs like a 130 or 140 or more to and then I just run that all day long. Again, checking it with breakfast. And and that's how we kind of get mostly normal days like I get ahead of it, I see the increase happen overnight, that shift sort of happens in that two to 4am timeframe, most of the time. And I just put an override on the compensate because her basal needs went up, but I don't think it's going to stick around for the next few weeks. And it usually lasts a day or two, sometimes a week if she's actually sick. But I always use that overnight as a measure to say how much more or less does she need. And use an override and I don't really mess with my base settings often because they usually work occasionally, if you get up to like 150% override, because she's sick. Sometimes the carb ratio has to be weakened a little bit because doesn't scale up quite perfectly. But I'd say 9090 plus percent of the time, it's just a 1020 30% override for the day, and then the day looks normal again. And then I watched the numbers overnight and wait for it to shift again. And then I cancel it and then we go back to normal settings. And it's it's really kind of magical to be able to see the patterns. Once you get some consistency. You can see those patterns coming and get ahead of them. Yeah,
Scott Benner 1:07:03
well, listen, I can't tell you how thrilled I am that you found the podcast and and reached out to me because I think that I listened, I edit these shows, obviously. And then I listened to them for sound. So I end up hearing them two or three times. But I'm gonna listen to this one dozen times, because this was a terrific conversation. I just I can't thank you enough. And can I can I ask you, right here, start thinking about how we could take a person who knows nothing about any of this? How do you explain this to them? Like that's the thing. That's next, right? Because you and I are involved. You know, and everyone listening, you know, is involved with their kids, hopefully, but everyone wants to not be and they don't want to burden their children with teaching them all of this like is, is an algorithm based insulin pump ever going to be? You know, I guess skinned in an app where you don't need to understand what's happening behind the pretty picture on the front. Like I hope so.
Kenny Fox 1:08:03
I think the hard part right now is the settings have to be like that's why I think control IQ seems to be working so well compared to like a Medtronic where it's trying to figure out your settings. But then control IQ just trust your settings. So you have to if you can get it close, then you'll get good results. Just like loop if you can get it close, you'll get safe and good results. But I think the real trick will be how can could that Medtronic idea of just figuring it out for you ever work? I'd love it. I'd love for that to be true.
Scott Benner 1:08:30
Yeah. Okay. All right. Okay, I'm gonna ask you to hold on for one second. I'm a thank you first and, and I'm gonna ask you because I'm gonna ask you a question. I'm not gonna let anybody hear the answer to I just like the screw with the episodes. So. Alright, guys, Kenny's done, but I'm not done with Kenny. I think it's pretty fair to say that Kenny will be back on the podcast at some point. Kenny could be Jenny loopy. loopy Jenny. Kenny could be loopy Jenny. Kenny Jenny. Kenny loopy Jenny Jenny loopy Kenny. Kenny Jenny late. I'll work on it let you know. huge thank you to Dexcom the Contour Next One blood glucose meter and touched by type one. Please go to touch by type one.org dexcom.com forward slash juicebox or Contour Next one.com. To find out more about the sponsors. Continued gratitude to the community that has put so much time and effort into the DIY loop. Talk to you again soon.
Please support the sponsors
The Juicebox Podcast is a free show, but if you'd like to support the podcast directly, you can make a gift here. Recent donations were used to pay for podcast hosting fees. Thank you to all who have sent 5, 10 and 20 dollars!