35: CGM Series: The importance of CGM Accuracy and Study Design with Professor Othmar Moser

Episode 35: CGM Accuracy, Evidence Gaps and What the Data Really Shows with Professor Othmar Moser

In this episode, John Pemberton speaks with Professor Othmar Moser about CGM accuracy and the evidence gap shaping current guidance in type 1 diabetes.

Recorded at ATTD in Barcelona, this episode explores CGM accuracy, study design, and interpretation gaps.

Key topics discussed  

• What CGM accuracy really means  
• Why study design matters  
• Regulatory vs real-world gaps  
• Language and interpretation  
• When finger prick testing is required  

Guest : Professor Othmar Moser  

Links  

• DSN Forum + GNL CGM Charts: https://www.diabetesspecialistnurseforumuk.co.uk/new-cgm-comparison-chart  
• Full show page: https://theglucoseneverlies.com/episode-35-cgm-accuracy-study-design/

Disclaimer

This podcast is for education and informational purposes only. It does not constitute medical advice and is not a substitute for individualised care.

The Glucose Never Lies® is independent by design

We do not accept sponsorships and advertising. We operate via education grants and donations from listeners who value independence. So, consider:
Buying the GNL a Coffee:
https://www.buymeacoffee.com/jspfree2

Enquiries

Collaboration: John Pemberton — john@theglucoseneverlies.com

Creatives: Anjanee Kohli — anj@theglucoseneverlies.com

Follow The Glucose Never Lies®

Website: https://theglucoseneverlies.com/

Instagram: https://www.instagram.com/theglucoseneverlies

LinkedIn: https://www.linkedin.com/company/theglucoseneverlies

LinkedIn: John Pemberton: https://www.linkedin.com/in/john-pemberton-587104361/

X: https://twitter.com/GlucoseNLies

Disclaimer
This content is for informational purposes only and does not constitute medical advice.

© The Glucose Never Lies Ltd. All rights reserved.

Transcript

SPEAKER_00 0:09

Welcome to the Glucose Never Lies podcast. This is from live from the ATTD in Barcelona. I'm joined with my good friend, um, scientific advisor for the Glucose Never Lies and mostly partly all around good bloke, Professor Off Momosa. Hi, nice to seeing you again.

SPEAKER_01 0:24

Um, so it wasn't too long ago, a year ago, in ATTD in I can't remember it was I think it was ESC in Vienna, and I only do remember because I'm Austrian, so I know it was Vienna.

SPEAKER_00 0:35

Yeah, and um obviously we did the podcast. The room's a little bit better this time. We've gone up in the world, we've gone from a hotel room into an apartment. Um, but one thing we did meet last year, we talked a lot about obviously exercise, which we might cover a little bit later on. But what we did discuss there, the CGM accuracy piece really started to kick off at EASD last year, and it became very apparent when we were talking to people about it. The language that's used in trying to explain that you need a strong study design and see the data is really difficult because you just anticipate if something's got a C marking, it should have data.

SPEAKER_01 1:14

So I I think from from my perspective, for someone living with type 1 diabetes and using CGM from the very first day, you know, with the scanning devices, then the integration within the HCL, now the integration within the fully closed up. I think it's about the main message is transparency. So we we need to understand as HCDs and at the same time as people living with type 1 diabetes, what is happening with my CGM drays in debt or uh another situation, and is it comparable to my, for example, black gilcometer? You know, this is what I do see. I do a finger brick, I have a kilcometer, and I want to know when I'm exercising, for example. It's always always under or overestimating. Oh, when I have an pizza. I would expect the CGM is always elevated a little bit. But I think it's about um having all the data available. Um we we're publishing and the companies were publishing, and then we are living in a good world. I think it's CGM, all of them are good when compared against 20 years. Um we must be totally clear. But the knowledge where a small Achilles hill improve the life of us as researchers of the companies and at the same time of the people living with type 1 diabetes and type 2 diabetes.

SPEAKER_00 2:39

Yeah, I mean, I absolutely agree with that. And we sort of sat down and said, okay, if we're gonna get a strong clinical voice out there, obviously you're very well known in the space. I've got a lot of time and energy that can be funneled in the right way, can really kind of make things happen quickly. But then we sort of said, well, we need to find out the opinions of the people who are leading the major organizations. So obviously you reached out to the ASD side, to Chantel, and obviously you know Tade very well. I reached out to some of my pediatric colleagues and also some of the people I've worked as on previous CSGM accuracy paper, and everyone was interested in the concept because they were having the same challenges trying to explain to their colleagues who are dialotologists, like what is all this study design and transparency? So we kind of knew we had something. And I the bit that makes me laugh is I looked at first draft and we've got a both of a looking out. It's really a bit of word, but it's good to go out to the first set of people. And we did it like a modified Delphi where we said we just didn't set say who else was in, just sent it out to the people who we thought would be really useful, some people blind, but a lot of people came back positive, but it was really interesting without that group think, there is a lot of good stuff that came out. I don't think it would have come out if it would have been in a group could have poshed.

SPEAKER_01 3:55

I I do expect the same, and this is what we see so often on these so-called advisory boards when we are discussing novel technologies and all the stuff. So if one has a very strong opinion that other people are following, so I do see that this blinding in this process of setting up the paper, and maybe you you have set it up. So um I was just evolved to be honest. So you you are the the brain.

SPEAKER_00 4:21

Um you get 20 back and then they're all saying something slightly different, and you go from the 1700-word thing to then up to 5,000, and then you overdo it, and then you send it back out, and you're like, no, no, it needs to go back from 5,000 down to 3,000, please with some tables. But over a few drafts, what became clear is three concepts which you decide. Yeah, first of all, we need transparency data, which we'll dig into.

SPEAKER_01 4:43

Which is from my perspective, the most important part. It can be, it can be a very accurate CGM, it can be, you know, not so accurate, but as long as there is the knowledge where is the Achilles here within the 24 hours line, it's totally fine from my perspective.

SPEAKER_00 5:00

Well, you take that from here, and I think that's really worth shilling on because we are going to get into the weeds about study design, and we're going to get into the weeds about calibration alignment, and there only exists that data for three sensors, three or four sensors. We have knowledge, deep knowledge on them. And you go to it and walk on the the tech lab ACTD today, there is about 15 CGM companies. Yes. And obviously, I'm a pain in the ass with them because I'll just look and go, wow, you've got a mod of 6.7%, and it was 21 days this census lasting. That sounds amazing. I'd love that. Do you mind just showing me where that data came from from the 6.7? Oh, it's data on file. You have to come into the medical booth and with like sideway your life. And then when you get there, they give you a little card that has got a scrap of data with no study design, no information about the patient. It's just, we had 3,000 paired readings, and this is it's like it may be a very good sensor, but without having visibility of the data, it's an unknown risk. And when we've got good devices out there where we know the risk, it seems uh incomprehensible to actually take that risk.

SPEAKER_01 6:12

And again, to be honest, I prefer using a CGM with a mark of 18 or 19, but when I exactly do know where the system is struggling, I would prefer it over a CGM with a mark of 3%, but you do not know. Is there any issue? Are the data really that accurately performed, which we all would assume? So again, transparency, this is so important. And then I think everyone is everyone is happy from all the sides. I think this is what we did also see during ATTD already at some posters, at some oral presentations, um, companies and also researchers are more open out about the data. And this is a very good initiative. It goes back to your paper. I'm quite sure also to the other papers that have been published in a similar way, that the data are now fully available. This makes life easy for all of us.

SPEAKER_00 7:10

Yeah, and absolutely. I mean, when I first did the uh C marking paper back in 2022, 23, you couldn't get the data on fire from a lot of the things. At least this time now, what we've done with the DSM forum in the UK is we don't quite have a we've got ICGM, we've got some recommendations for the IFCC, but nothing is like set in stone for everybody. Yeah. So we just pulled together what are the five or six really important questions to ask and give each CGM their study a score out of five. If I only done that in 2022, I'd only be able to fill out three of them. Yeah. Where now it's kind of like can fill out ten. And of those five have got a marker of five or above. So for others, we're like, well, if five have got a solid design score of five where they've got lots of data, where they've got it tested reasonably well and type and diabetes, etc., then it makes sense. We know the risk of those. Why would we compromise when there are five good options on something when you're talking about insulin dosing and you're only taking risks of people's lives? So we've developed that, which we'll drop in the show notes, which kind of puts us into the realm of if you're going to test a CGM and you're going to want to use it for insulin dosing, you're ideally going to want to stress test that sensor in the potential hardest day it will have, because if it performs well in there and it's reliable and usable for insulin dosing, then all the other days could. But if you don't test that day, it may be good, but you don't know. And I think that's the point now where we've got transparency of data's quite decent and now loads of level two, which is how do we make sure that the CGMs are tested in an effective way? So where do you see that situation at the moment?

SPEAKER_01 8:55

So let's focus on a stress test now. Um, there are two very nice stress sources. One is a meal challenge, one might be an exercise challenge. Um, I think for the meal challenge, it's even quite easy because the meal challenge would always be associated with an increase in glucose, even if it's high protein, you might go up a little bit. So, what we do not see is a decrease when you consume carbs. But this is still a stress test. For example, play a little bit with the insulin dose when someone would forget. I am unsure if this should be a general, a general recommendation to test it, but a stress test based on meals should be included. For example, we have a study ongoing now with kids, and they have a pizza challenge, a spaghetti challenge, and uh, what is it? Um I think further rice, curry, or whatever challenge. Okay. You do you also don't eat any of those. I'm filling in. But you know, it's it's fun, first of all, for the kids in part of the trial, but we see also this hyperglycemic values. We can assess, we can assess the peaks, we can assess with a very high rate of change eat glucose when it's dropping. Same for exercise, for exercise is quite diverse. We see different responses. There might be an increase, they have a decrease, but from my perspective, it should also be included. And why is quite clear? We do recommend a physically active lifestyle to people with type one. And then we should test all devices also to a physically active lifestyle. This holds also true for artificial functors, for CGM, and again, uh it's about supporting people with type one, supporting THCP, and also supporting the companies that they are receiving further data to improve algorithms of CGM, algorithms of automatic dosing of insulin. So it's a win-win-win. And this is, I think, what we are all angry for.

SPEAKER_00 10:54

I mean, I'm definitely going to come to your two recent publications where your PhD student Serena has done great work and really drew out something which was um, I guess it just corroborated some of the other things that we'll get to, but it's really nice to see it done exercise personally, because as you mentioned there, you you did the paper back in 2020 where you showed you know accuracy does drop off during exercise, yeah. And we just got a lab and recommendations. So we'll get to that, but absolutely. So stress test, and this is why I've got a DSN forum chart where you get studying at the score of five. If I was a manufacturer, I could gain that. What I would do, because you have to have a percentage of redoes 8% in the very high, sorry, 5% in the very high, and 8% in the very low. If I knew that, I could just give manufacturer and go, I'm gonna go into an investigation, and this is how my protocol bring them in one day with a glucose level of like uh 200 or you know, 11, give them a little bit of carbs, crushing into the 1617, nice as well. Leave them there for eight hours to protect your values. Next time you bring them in, start with them at about four, you know, five or a hundred million, a little bit of some drop down. So you get the coverage, but you don't get what happens in real life. In contrast, the the publication by the IFCC team have suggested the best way to stress tests to get everything covered, both the past rises, the fast falls, the highs and the lows, yeah, is the meal maintenance charge on the same day. So six grams of carbs, yeah, push it up. When you get high, it's big instantly drop down. Yeah, you therefore cover all of the range, and you also put it in the hardest hardest day of the half. And then if we know it's accurate then, the rest is cool.

SPEAKER_01 12:27

I and think to be honest, this is this is real life situation. I am quite aware that we have also other opportunities, but it's not um um performable globally. And so we have in Graz, for example, hyperglycemic um clan procedures, so we could do it also artificially, but this is not what it is about. It's not so we had also uh requests to do CGM um assessments during hypoglycemic clan procedures, so you fix them on a 3.9, on a 3.0, on a 2.5, and then let them go up again. But from my perspective, it should be extremely practical. So maybe the 60 gram carbs challenge is a proper way that reflects real life situation and having this. So maybe it's not just solar that you were correctly saying about highs and nose. I think you know, one driver for performance struggles is the way of changing things. We all know, but at the same time, we must be clear how often do you blood glucose assessment from fingerboard, exactly when you are hiring. So it's a further safety tool to do blood glucose assessment, but at the same time, from you know, receiving approval and assessing MART and error grid analysis and weight of change dependencies. Try everything to include in the study protocol. This is always my impression. And we tried quite often doing this. So I think we have over over 20 publications, mainly within exercise, around exercise and CGM performance. And it's it's it's of interest and should be of interest for all of us today.

SPEAKER_00 14:09

Well, it's just to check my glucose. You go for it. But it's at the stay of life and you're checking your glucose now, like you're making that decision based off that CGM value. Whatever you're doing now, you're deciding whether you're okay. Do I need some carbs? Do I, you know, give yourself a bit more insane, whatever. You're your next 20 to 30 to 40 to two hours is based off what you've just done off that basis. So you need to have a level of confidence that it's never going to be perfect, but it's going to be accurate enough to make decent decisions. The vast, vast majority of this. Consider this. If you go to a car dealership and you're offered two cars and you're told that they both drive maximum 80 miles an hour on the um motorway, and they're both eagle cars, one is cheaper or half the price, but the rest is in. Then he said, Okay, we had a safety test. And one car has had you read the documents, it said, we did a crash dummy test into a wall at 10 miles an hour and it was fine. Yeah. And you have a look at the second one, and he said, We crash tested 18 miles an hour into a wall and it was fine. Yeah. Now the both cars could be equal, but which car are you happy to get into and take your kids into?

SPEAKER_02 15:17

Yeah.

SPEAKER_00 15:17

So it's first of all, depending where I would be living and how much money I do.

SPEAKER_01 15:21

What would you have? I want to have a safe and cheap option if I could choose. And if the safe option is more expensive, I would pay from my perspective for the more expensive option, but I want to be safe. Um, I think in general, we must declare all CGMs there that are approved are safe. Yeah. Um compare it always against what was 20 years ago happening with CGMs. So we were happy if it might be a lot of people. I remember the City. It was like a javelin. Yes. So we should be happy. So think about it, about your diagnosis, how it was like, how it was my took.

SPEAKER_00 15:57

So basically, all the ones that I talked about. So you've got the Roche Smart Guy, you've got all the FreeSty Libra products, you've got the Dexcom products, and you've got the Medtronic products. There's ones with your design score of five out of five. They're the ones personally, I do it ones. Personally, I would support the people who I was going at and say the data is sufficient enough to understand the risk. And then you understand the risk part is the 2020 and the four to fourteen. This is going to come to think freaking. So the easy way of thinking it, I'm not going to build up 2020 as you can look it up. But essentially, if the whatever percentage you're reading is in 2020, you're not going to make a mistake. A good clinical decision. That doesn't mean it's onto the fingerprint, but it generally means it's within one and a half, two. You're not going to make a mistake. Yeah. And then the second metric is 4 to 40, which is a wider metric. And basically, the percentage of readings that fall outside of that are the ones that will definitely lead to an instant dosing error. Either too much insulin or you'll lower and maybe pause the insulin system. And here's the key part, and this is why I want to bring you in. Yes, even the best ICGM sensors, the best sensors in the world, have between one and 0.5% of those readings. That's like that. How many readings a day do they do? 200 and something, something like that. 285 over a month period. And you're going to have guaranteed at least five or six occasions over a month where you don't be outside of that. And the key point here is that's amazing that it's only those number of readings. And the CGMs are so good. But those five or six readings are the critical point where you're making a potentially life and death situation decision, or a system is making it for you based on that value.

SPEAKER_01 17:41

And you need to have yes, yes, and no. So the issue is for sure there's a clear, so we have this clear assumption wrong reading, wrong decision could be fatal.

SPEAKER_00 17:55

I mean the thing is, I'm talking about that's like the percentage chance that that is extremely minute.

SPEAKER_01 18:01

However, when there's a chance, and we do not have at the moment these data. So we must be clear. So there was there's no evidence out um that's saying this one percent of out of the space meetings that something happened. It's also difficult collecting this data.

SPEAKER_00 18:18

Okay, I'll say something I'll say something there as well. When when I'm I'm presenting there, like there is a worst case scenario of people die on diabetes all the time.

SPEAKER_02 18:25

Yeah.

SPEAKER_00 18:26

And with CGN compared to theme pixel they had before, that number will have gone down. It's massively improved. So I'm not in any way saying, oh, do you know what it's worse than it was? The baseline case would have been the number of deaths would have been off the chart. It's gone down massively because all the functions. But when you and I live with diabetes, you know it's still not a non-zero chance that if you have the unfortunate thing of a collection of readings over a period of time that are outside that 40-40 for whatever reason, there is a minuscule, minuscule chance that it would lead to a very tricky situation if you haven't been educated and given the tools how to recognize and what to do. But when there's a non-zero chance, you always have to apply, of caution of principle, and I'm sure you have with you, and I have with me, something that means to go when you get that situation going. Just doesn't smell right to me. What should I do just to check? You're about to tell me you were taught back. I was taught back.

SPEAKER_01 19:26

I was I was I was educated that it wasn't exactly that point. So, first of all, I had no CGM on my diagnosis, so I had a full understanding of blood glucose measurements. And I think this is why I still do on some occasions a blood glucose measurement. First of all, I'm interested in it. Yeah, this is my personal reason, but I do see just for me sometimes that when I'm in exercise and I have already symptoms, that there is this lack time. This is already known. This is what I do. Like what I wanted to say before, we see more and more young stars with type one, they have the diagnosis, they never do black pickles assessment. So they are so confident and just and now comes a critical mind. I think you and myself are idiots because I'm quite sure if our diagnosis would have been five years ago, we would have maybe a completely different mindset for our own diabetes, for just one and we wouldn't do any black people. This is what I would expect. Do you think we do I mean you're not too much? So we're doing blood reconcilation, so it's part of my diabetes history. But I'm quite sure if my diagnosis would have been five years ago. I think they do maybe one or two blood precursors within months.

SPEAKER_00 20:37

And I think I have no clues, but I thought I don't know. I think different because I reckon for you, your amount of detail, you would have gone in, and this is a key point here for you all. You would have gone in and looked at the accuracy of the good fingerprint is the I said ones. They have over 95% of readings in 1515. And I just told you that they're CGM that are on 90% 2020. So they've got 1550 is even more accurate, and they're even higher at 95%, and a lot of them at 99%. That basically tells you almost, and then with non-outside 4040, non-outside 440. So you'd never have that problem with a finger unless you like you do something, as long as sound was good.

SPEAKER_01 21:16

So I think, and I think this is a major, uh, very important information. This is what we are still teaching. Always take care that there are no sugars on your finger when they do, for example, actually or whatever, wash your hands in advance. And the thing I'm always afraid of is So I'm doing it because I'm extremely interested in the CTM story for myself.

SPEAKER_00 21:36

But I'm and also you're checking the where is it, where is it lying in relation to my gold standard 95% above 1515, occasionally just checking in. Which sense do you use if you're on S?

SPEAKER_01 21:50

Um, I'm chopping around at the number between Epic So can be honest, I'm using an AID system and I only have two options. Sure, yeah, this is just text problem and it's But um and this is why I'm playing around. So I'm quite depending on the situation where in which sensor do you want to use? I also used um the Madronic sensors, I used uh the Sinal Care sensors in both chance. So I'm interested in it also, for example, for Sinalcare, it worked absolutely well for me. So what I wanted to say is what we also see quite often and have no real interpretation, that for one, for example, the Medronic sensors, it's really well working, and the epot is not working, and the dextrema is working in the middle, so we still see this individual I don't know, but is the I I can I can answer that question. Do not answer with calibration. No, no, no, no, no, so calibration. I'm not calibration by amazement. I mean I'm written calibration.

SPEAKER_00 22:49

No, no, not that either. It's not that. So I would be curious to know. So if you look, the best study for me is that that three senses study, although it's actually four people, the beauty of it is they they've got something called the CGD, which is uh a set of metrics which kind of it does the ICGM stuff, but what it does nicely, it puts this plot of the individual senses of where their bias is to the comparator. So if the comparator is venous glucose or capillary, whatever, you've got this line across the middle, which is what called capillary capillary, and across the 24 senses, if you start at the end of those are under-reading, you've got probably 15% reading way under and 15% reading way over, and in the middle you've got a nice selection. So you just think if you've done that and you can see the lines across from Medtronic, uh DexCon, and other, and you could see if this day I got the sense of the G6 was on the file on the left here, and I have the Medronic on the file on the right, it's expected over 24 centers, but more experiences like, oh, I don't buy out.

SPEAKER_01 23:52

And so I think this this is very also very important information. So I get so often asked which AID system are you using, which center are you using, and this is not the global answer that might be correct. So I choose the center that is exactly based, and that will be a thing with me accurate. Yeah, yeah. So it could be it could be the new one the instincts, it could be the G12 of Dextrone that might come up in five years. So it's an individual thing, and we are at the moment analyzing mainly median curves in our accuracy assessment. But from my perspective and my research group, we want to follow more showing in and table for each single thing that we do out the full data, because then you see where the median line, the interquartile range or the standard deviation is coming from. So what he learned when we have written this paper, where I was being being an out being part of this, uh, will show for all the future um accuracy data. So it's not just CGM and accuracy assessment, I will show each individual response in each situation in the round table. And if the journal does not allow me to include it as a main graph, it's in the supplementary material.

SPEAKER_00 25:11

Do you know the graphia? The graph that I love is that the same graph where it also reminds me this this will resonate with user exercise research, and if microveil is listening, death is listening, it definitely will read because I handed them for ages on this piece. Is they did a study of about 120, but so they did a uh a secondary analysis of four studies put together where they said the higher they start, the harder they fall. I think it wasn't it wasn't hits as far as with this graph showing the drops on the far left was those who didn't drop at all. On the far right was the greatest drop, and obviously the mean drop was about like two millimoles per meter, and you have from the little as a rise of two all the way to a drop of twenty. Yeah, and then but obviously if you just read the paper, you go, Well, an average many 45 minutes, the average drop's two. Well, the average drop of two is about right for about 30% of those people, but yeah, the message that goes out.

SPEAKER_01 26:07

And I think it's sorry, it's two. Do you think this is true for all aesthetics? Of course it is. So the reality of the mean holds true for around 30 up to 40 percent of people.

SPEAKER_00 26:15

Almost a ridiculously flat response, which never happens in the biologist.

SPEAKER_01 26:22

But uh I think it comes back to the first part transparency. Exactly.

SPEAKER_00 26:27

Showing what showing that's why I like the CGD with it, because you don't yeah, you do get all those metrics, but you get a visual ego. And I look at how it went. Holy crap, but it also when it comes to calibration, how much we're gonna get there, is I'm just gonna box this off and say for me personally, any of those five senses we've got off five, I can make them work for me. Yeah, I know. I'm happy with 90% 2020, and we also know there's a risk of 1%. So, massive point. If you're listening to this, then I was gonna say it's not just people with diabetes, we've now got three new nurses in the team. They've only known CGM. And what's they've been told, maybe by us, maybe by the manufacturers. It's kind of like these factory calibrating systems are absolutely I've been on it. I've been around them and as a stanza. What do you teach about glucose system? Oh, these factory calibration systems they're better. I was gonna say they're better than fingers like you've got one in there's five percent, 15-15, not sensor, seriously, and then you go, oh, what do you mean? It's like that's you're not even in the same ballpark. So it's it's if in doubt, get the glue glucose meter out, then it will be a case of a couple of times a month, and just for me to do it. And then you've got a choice. Yeah, if you've got a system where you can optionally calibrate, you can get your for me personally, that's why I choose your optionally calibrated one. I'm not saying which one it is because it doesn't matter, but when I'm in that situation, my psychological confidence in the next few hours is all about do I think it's back on what my fingerprints is. So if I need to be able to optionally calibrate, I don't have that period of things it says off, and that's just personal thing for me because I do fingerprints, but then I speak to the nurses who come on board. We don't really teach, but I may not know.

SPEAKER_01 28:06

Maybe but maybe the truth is in the middle. Maybe I think one side we are the threats, yeah. Often blood recurs measurements, yeah. Yeah, maybe the answer is in the middle. So what I would prefer for myself is sometimes I'm a little bit overreacting in regards to blood because measurements that I'm fully about now it doesn't because to be honest, there is always some variability, and it's not this worst case scenario. And I never did detect it for myself in this, I don't know, 30 years I have to have one or whatever. So maybe the answer is exactly in the middle. I think we still need to teach how to do a black glucose assessment, mainly based on symptoms. And if you see a glucose stress that makes no sense, that you are early dropping after a meal, that you feel confident afterwards, do a blood glucose assessment, but in general, let it row and focus on your symptoms. You know, I learned that I had I have five blood glucose measurements per day available at my diagnosis. And normally I would have been only three per day when I had my diagnosis. But I wasn't the national triathlon dim for ospheres, I was allowed to do five blood glucose measurements.

SPEAKER_00 29:15

Maybe about 60 now, does it? I mean she's about 60 years old, isn't it? We're about doing two or three.

SPEAKER_01 29:20

Um so but that but it it also did work, but it meant at the same time that HPM and C was higher. So we have the beauty that is called CGM, and I think we can we've been double crazy. We can fully trust, but we should trust. But if they are sync and you feel better doing the black because assessment, do do whatever fits best for you. I mean, this is my last.

SPEAKER_00 29:46

The summary for me of this is CGMs, if they're at somebody who's asked for a five out of five, they've got accuracy of at least 90% 2020 and less than 1% or half percent 40-40. They are amazing, the best thing ever personally made a career out of it, and you know, and love it. But a couple of times a month, you may need to do it. Whether you choose or not to is your choice. You should be taught that that's the case, and you should be taught two very important things. Number one, it should be an ISO standard cool glucose meter. It's important. Number two, the specific scenarios when you're kind of got symptoms for low and maybe don't know what's going on. It's in doubt, check and saying on the high side, like you even though yourself got you look at your CTRL and you go, Yeah, with what's done in the last three hours, and I should not be this number. And if you have that feeling, just do a thing that can check.

SPEAKER_01 30:33

And what we do also say is when you have, for example, so we all we sometimes this is a situation that someone has symptoms of hypoglycemia with a blood glucose reading of about 7.5 millimoliliters, and then there's more blood glucomene around, and then they do weights to consume carbs. That's where we're saying go in with carbs, instead, you're a little bit elevated, it doesn't manage those ways for a blood glucose version.

SPEAKER_00 30:56

I just think you're a bit pissed off later, but you're better than being completely hypo.

SPEAKER_01 31:00

Exactly. So being a little bit higher is better than being extremely low. So but I think in general, from my perspective, you can trust the CGM completely if you find the right CGM works for you. And if you have synters, choose to do the black records assessment if you want to, or go in immediately with some cops. So I think that everyone needs to be able to do that.

SPEAKER_00 31:23

I want to say this now, and this is for all manufacturers, is also they need to be careful with their messaging because if their messaging is the CGM is more accurate than an ISO standard fingerprints, then number one, that's categorically incorrect. And number two, if within their literature it doesn't state quite clearly that fingerpricks from an isometer are superior, you may need to occasionally do it, which they all say, yeah, but they're a little bit elusive the wording to whether that's the main players or the new ones. I've been down to them all and asked them what's their advice in their book that are not fingerprints. There's no mention that if you couldn't do that.

SPEAKER_01 31:59

I'm sure if there's still the symptomatic storyline, you should do a confirmative plaque. But I'm sure no a joke.

SPEAKER_00 32:05

Well, this has had only a junk system, like CE Martin just passed passing on a junk to like it. Sweet. So basically, manufacturers have a responsibility to put in their literature. Healthcare professionals in their teaching programs have a responsibility to teach it. If they if they choose not to, that's fine, but you have to cook yourself because it's a non-zero risk that there'd be bad event. And when that's the case, if you're thinking about it's you or your kid, you want to be taught that from the beginning. But all that being said, roll it up, these things are amazing. We're not saying go back to doing thing picks, we're saying go back to doing things picks once a month, but have the kit with you along with a key termometer. If the proverbial hits the fan, you know what to do, you're taught what to do, and you're taught the right stuff. And when you come to clinic once a year or twice a year, what we're starting to do now is going, can you show me a book glucose meter? And then you go, we haven't got one. Yeah, but this is okay. You don't give me one next time, you split one. Tell me how you do glucose test, and it's hilarious because you kind of get it out, you have the jaby finger, the squeeze in the end of like let's go on to basics.

SPEAKER_01 33:06

Always keep in mind we can rip it off. So it it it maybe not in the discussion of inaccuracy when I'm in a walk. When I walk through the door, when I walk into the door, and I'm losing my CGM. And then you're, for example, sitting on an airplane and you cannot change. So when I'm driving, especially, or when I'm far away from home, I always have my my my uh Fukumiya with me just for, you know, I don't want to go home and insert it in there for two hours, one hour, whatever.

SPEAKER_00 33:34

Oh, it's probably so rough today. So I want to kind yeah. I know we've had that point. It probably sounds like we're gonna, and to me, this is the situation. I feel like the education and the evolution of CGM has just gone half a step too fun. Like it's gone a hundred steps forward. I believe you see to go half a step back and say, well, glucose meters are still always gonna be more accurate. There will be those occasional times, let's teach it properly and let's be honest and about it. I don't think there will always be more accurate.

SPEAKER_01 34:00

So I do expect the diffusion. Yeah. I'm I I do I I do I do think maybe it's not mentioned in the social space. Well, yeah, if it's good. Yeah, if it's in the space, but what I do see, and I'm quite sure that the companies might achieve, let's say, the mark because it's the most common below the 5%. I'm quite sure we as we are sitting here will see a CGM and will worry CGM maybe what next year, but maybe in 20 years. I'm at the SQ on that. I'm the best industry that you do each other enzymatic. The companies have are fully aware this that this this so they always even if we're achieving marks of of you know that being below 20, below 15, below 10%, um, I think they are still working to improve.

SPEAKER_00 34:53

Final one on this piece, because I know you uh this you said this exactly the same to me. If I wanted to prove that during exercise you need to have a finger cricket with you, I can do it right now. I push my glucose to 12, I give myself a whacking dose of instinct, yeah, I'd wait for it to kick in for about half an hour, and then I would go for a run for 15 minutes. And I bet you when I got back, my finger cricket would be two and my CGM will be about nine. I'll give it worse times. It's because when it drops so rapidly like that, the lag time, as you've shown, increases from two minutes to 15 minutes. Yeah. And that it's not like it's not fault on the CGM, it's just changing physiology.

SPEAKER_01 35:31

So respect during exercise and they drop it in the technological nectar time, it's a physiological next time, and this is normal glucose. But as you were saying, when people are where then we treat array, and this was why we defined the threshold of seven minimolboliter that early with carbs to overcome just physiological nectar.

SPEAKER_00 35:50

So, Sabrina, I'm sorry about this, but this is your next study. I would love you to do a CGM accuracy on exercise. Yeah, you start the glucose level at 12 with no incymbols on board, and you start with 0.15 units per kilo or not two units per kilo an hour ago, and you measure the accuracy over a 20-minute, 20-minute exercise. I guarantee you the MAR on the first one will be about 10%. On the second one, it will be about 40%. Yeah, and it just shows hiding sun on board, rapid movement of glucose is the time. That's why you set your CGM target hydro exercise, and it's why you always have your finger prick with you because you've also got sensor losses and stuff. Epicides is so essential. Well, I think we've done that to depth. Yeah. But I hope that's a message that resonates with people, especially if you're a new listener and you have never done a finger pricking. Like, what's the point? There is a point, it's a rare event which has catastrophic potentially catastrophic things. You need the ability just to wipe it away. And if you haven't got glucose meter and you know what to do, it's an it's a no event for you. So write it off. So we've got the stress test, the IFCC did it. And what we saw is when you do a stress test, something becomes very paramount. These CGMs are taught by our data set how to trace the glucose, how to follow it in a training data set. Now, depending on how that training is done and depending on how the algorithm responds to fast moving glucose, will depend on how well it tracks in real life. And they will all have something different because they're all proprietary. You don't know what they are, they all use different data sets, and they all have different mindsets. So, for example, if I was a company, one of the first companies to come out with the AID system, do you want a sensor that tracks right on the glucose, bearing in mind what I just told you before? But over a batch of 24 sensors, some are going to under-report, some are going to over-report. If you get one of those over-reporters and you're in the top end of capillary glucose, you're going to be over-delivering, you would destroy the field of AID therapy. So it makes sense for the early players to underdo the glucose to prevent over-delivery, which is a safety-minded, very sensitive decision. But then as the sensors progress, potentially go, maybe they don't need to be as safe as that, which can be in the actual space, or they've got a better algorithm because their engineers have figured out a way. But what you did see in that study is when they put the stress test on, two sensors, and we're not going to mention names, just no point, two sensors tracked between capillary glucose and venous glucose, which goes up by the difference is 30%. They tracked right in the middle and back down one sensor red on me. Yeah. And that sensor is absolutely a great sensor. It's safe, you'll never have different injury. But when you looked at the data, that sensor reported for the same people 10% more time in damage than the other one. And that's not bad or good. It's the same, it's a decision made by whoever the engineers that when glucose is rising like that, be conservative because we don't want to go with it.

SPEAKER_01 38:52

And for so again, for me as a patient, and you you were talking also about the history, I think, or when I would have a child with talk on my quite safe with the safe solution. Yeah, yeah, yeah. With this, with this little with this little under reporting, but at the same time, you would correct the state. If you have the overall assessment, this has for the for the three-month period of assessment of time and range time interviews, it has influence results. So maybe again the truth might be in the middle.

SPEAKER_00 39:26

But you know, the beauty of that study is it revealed that actually the calibration algorithms that these things have for the three popular ones that we know are quite different. What about all these other ones that we've got no data? They could be reading above Capillion. That's why you can't take the risks because you don't know. Those three I know exactly. And this comes to the point because then when you then we know in that one study, over reports high range by Tempeston, and then you did an exercise for discipline, and you found one.

SPEAKER_01 39:57

So I think we found in general very similar results in almost imperative use. It was not really hardcore exercise we did, but we did something like yoga, we did multiple. Yeah, yeah, sure. And and it's for in exercise, the similar situation as under, you know, this resting uh situation. We had their hotel settings, so all the participants were sleeping in a hotel, they were constantly with us. And you know, these this data has been confirmed. Um again, all the CGMs used were in general at the rate. It must be safe, they're fully approved. Um but I think what we did see after performing the study that all the individuals, the participants, wanted to see their individual data. And what we did see was that they chose then a different sensor, but some also moved to this near Vioca. Some of them, for example, moved to the so-called less accurate sensor because for them they were accurate. Oh, yeah, yeah. This was was a was a thing I was not expecting at that extent.

SPEAKER_00 41:12

And did the people who were warned that the sense that it was underreporting the same as that, did they choose the one that did they decide to move the one that actually went closer to the actual glucose physiological space, or did they stay with the safe self-interest on average?

SPEAKER_01 41:25

So they did they did not. I could not answer. So they were saying, um, show me the data, and then they received the data for it seven bar. The absolute difference is you've got you've got Serena, you've got to follow you've got to go and find out those in the issue. So we did we did we did send it to them simply and then we're afterwards when we received the email a topic a couple weeks ago. Okay, I moved to this or to that center, but it was not just that all of them moved to one. So we had all the options, and we did not I'd love to know that we did not expect that.

SPEAKER_00 41:56

I would love to know the ones, what what they were running.

SPEAKER_01 41:59

The reasons, the reasons we didn't ask, we were thinking Sabrina, Sabrina, get on the case.

SPEAKER_00 42:04

We're gonna find that information. What we're saying here is, and this is backed up because this was in the clinical position paper, is what you've shown, what Guido shown in his study, yeah. You showed him your exercise study, is that some read in the physiological space between capillary and venous, and some read underneath, likely because of calibration things. And what that means is there's a difference in timing range for the same individuals present glucose. And then you look at the AID literature, and what you find similarly is that AI AIDS system that uses that sensor to under-report glucose and over-reports timing range, they have 75-76% timing range action with an A1C of 7% or 52%. And then for some of the other sensors that use those ones that measured in the space had a timing range of what 65% and are also HP and C. So the messaging is clear. They're both safe, they're both accurate, they both get the same HGAC, they just have different timing ranges, and that's not a problem unless you know that. So you know that, and I know that, and I know the people who are using that system that over timing range. You come back to my clinic and you put 70% in now, unfortunately, you don't get the pat on the back because you're fully done. I'm optimizing your settings because I know you need to get to 75 to get the same. Yes. Joe Bobs, who comes in a different system with a sense that he's in a space that doesn't report. Yeah, I mean, I know that. So none of my patients are disadvantaged. But because an international target's a 70% tight and range, most clinicians, once you come in 70% and be like, you've heard this on, I'm not gonna mess with you anymore. I'm just gonna, and that for me is a bit unfortunate. And it's also unfortunate if people see that 75% in the literature or on Washington magazines and go, that's the 70th because that gets more time in range, and that's actually not going to convert. They do get more timing range, so the market is correct, but they don't get the future benefit. And so I think as these stress tests come out, first thing we'll notice, as we've seen, is this big difference. But then I think once that becomes Clear what are the manufacturers going to do? They're going to align to write in the middle, they're going to change the algorithmic algorithms.

SPEAKER_01 44:08

And I think this is this is the good thing on doing research. And you know, this study from GUIDA was massively discussed over the in all the different ways in positive and negative ways. But what is happening, companies and researchers are still improving. Yeah. And this is what it is about. So let me say something very clear. I'm not happy with a mark of seven percent. But this is what I was saying. So and this is what we do see that companies are chasing it massively. And I think this is what I was saying 20 years, mart of 5% for all the devices, maybe. Um and at the same time, it's it's a company. Companies need to to sell, to have proper numbers, yeah, but at the same time to have money to do research. So it's a it's a cycle. I would not call it a wish a cycle, but it's a cycling. So I think it's for a good thing out of the extremes and studies, yeah. Because now everyone is again improving. Yeah, exactly. I don't want to be at the final stage and constantly want to improve devices.

SPEAKER_00 45:11

I thought if the manufacturer will be listening to this, who sort of uh who underreports the glucose and over reports time range compared to the others, you know, kind of listen stuff. These guys are kind of saying that uh, I think no, they get exactly the same H and C. So, in terms of what's gonna happen to you in the future, they're all as good as each other. You just can't claim superiority. But what you could do, you could probably claim you get the same HP and C, but reality is we're probably safer because you're never gonna know what would deliver the incident because the individual thing. If I was looking for the company, I would be thinking about that.

SPEAKER_01 45:43

We did spot so over underestimation, not a question comes in detail in which situation is you're doing what, and then everyone is improving.

SPEAKER_00 45:52

I can't wait to see the AID system that has now got one of these sensors that underreports, yeah, and now they've got an option for the other sensitive reading is data.

SPEAKER_01 46:04

And this is this is what this is what it is about. Now I can choose. Yeah, I think to be honest, so if I would be the boss of all the big companies for AIDs and CLCGM, I would be forcing everyone that for every insulin pump I can choose whatever sense I want to do.

SPEAKER_00 46:24

Yeah, and what actually hopefully will show or over time as data comes out, it appears from what we've seen here at ATTD, is that they the both sensors get the same timing range, which means that the algorithm's good, because in theory, the sensor that reads in the physiological space is going to be good at higher 18, so it's actually able to keep up with it and attain the same timing range. My guess, and I'm happy to come on here whenever the data published and go absolutely wrong, would be if you follow those for 90 days, you just switch to the 90 days on the sensor under report in theory, and 90 days on the sensor report, you've got the if the algorithm's good as it appears to be, you're gonna end up with the one with the reports in the space having a 0.5%. Then that company can go, we have superior outcomes. And that for me would be great because if you've got a clear winner on time and range, if that's what you want, because you choose that. Yeah, that's the 5%. But I will I'll I can't wait for that study to come. I'm happy to be wrong and go, do you know what that's happened and that's science? Yeah, it's good, it's good for it. So just to summarize that, we've getting there towards transparency, but we need it clearer on CE marking. FDA is absolutely fine, but we need to be driving. If you're a healthcare professional or a person with diabetes, I think when you go to ATDD or you see someone, you ask that question whether well, you can go to the DSM forum chat and find that some of the design schools have done it for you. So there's a simple thing: don't accept MR numbers, number one. Number two, we ideally need stress tests to make sure you have the worst awesome day, it's covered. And then also that reveals where systems are reading, whether that over. And this is a new concept that potentially means for some of the sensors you'll need a little bit more time in range to get the same HVMC. But it's easy to get back to time in range because it's not going up as fast as so it's not like it's a disadvantage, yeah. It's just knowing that when you go back to your clinician or yourself, that if you're one of those sensors that's having that reads, you just need five more five percent more to put it's gonna be easy to do. Yeah, so it's not a good or bad thing, but it's even safer. No problems with that. So, but once we get clearer on that piece and we know which sensors are there, people go on with glucose, it's actually documented where sensors read and what you might want to consider. Um but ultimately from a manufacturer's point of view, I'll say this is I did a webinar for Insular and we discussed that 70% time isn't equal. And the whole purpose of that was obviously for them, they wanted to show that there was no difference in terms of hateful things, which the systematic review showed. So actually, from a long-term perspective, and it doesn't mean everyone's better or worse, but the point of doing it for me was to say the people who are listening to are using the other product is to the questions to listen and go, oh, so if you come in on that, they're probably gonna need 75%. So if I've helped 3,000 people via a clinician to go, we'll just remember that is a massive win because it's just knowledge, and there's nothing, there's no good or bad, as I said. I think it should be easy to get it because it's about it's about going.

SPEAKER_01 49:28

Because then I think it's fine. And I I think again we are all on the same route improvements. Yeah, it's never stopped to.

SPEAKER_00 49:35

Yeah, absolutely. Yeah, so on improvements and on changing perspectives and on what we once believe to be the best advice, yeah. Again, it's kind of like it's not it's not painful. I think it's great that you're that you're actually and more than willing to dig in, ask the questions, then go back. Yeah, do you know what? I think it's gonna be the case. So I'll leave this and say, I was fortunate to be on the position paper there for AID and exercise. Yeah, really enjoyed it, saw how much hard work yourself and Desi put in. Um I you know, random comments coming in all over the place. And I'm sure everyone else is a little bit more organized than me. But having to deal with all that and pull it in, pull a consensus on when you're on a LA therapy, these are the things that will stop doing, though, these are the drivers. And then you go and do a half a million data close to half a million. So what things have changed from what tell us a little bit about because that half million makes my eyes water. How have you got to assess it?

SPEAKER_01 50:32

You're some data scientist somewhere who's doing yes. So I think the storyline is which which comes back to the to the word improve. So we have we have we have exactly we have published um you know 2020's position paper for exercise and diabetes. We did the same for AID focus in 2025. And there are clear statements where should be the focus on and when to avoid hypokarcemia in the word hypercysemia. So we have together with with a big company um where they were assessing over years CGM value of the request, can we assess particularly exercise? And we did also deliver their constantly data. There are also a whole lot of participants from our side included. And yeah, at the end of the day, I think it was it was um data from 3,000, close 3,300 people with 1 diabetes, close to half a million exercise sessions. It also includes intensively walking, not just exercise, but we were also able to distinguish between running, walking, blame, then it's vain foot. So amazing. So and and and the major learning was first of all, there's no real increased risk of during or in exercise hypoglycemia. The major risk of exercise hypoglycemia is the first hour after the exercise session. Hypo. Hypo, sorry, for low readings. In general, when you do morning exercise, you do not see an elevated risk of hypoglycemia at all. However, when you do evening exercise, you have a massive increased risk of hypoglycemia at around 3, 4 a.m. in the morning. So we were discussing even on AID. Even on so it included all the hypothermies. So it's everything together. And you know, we were 2020 only discussing or mainly discussing what to do in exercise to avoid hyper. We were 2025 only discussing what to do to avoid typos.

SPEAKER_00 52:45

Exactly.

SPEAKER_01 52:46

But this is what it is about. Maybe we are idiots. Maybe we could we should focus, you know. Desi, Mike, you, myself, and all the people should focus more on the post-exercise management instead of solely focusing on the in-exercise management. But at the same time, I just think we have not all the rules available when you have type one, you know, correct one and then it's fine. Um, but focus on the post-exercise phase. And the first hour looked so important, and maybe I would change one sentence in our position's paper where we were saying that the end. This is very important. At the end of the exercise session, go back to the regular glucose target. I do not recommend this anymore. Stay for two hours on the elevated glucose target.

SPEAKER_00 53:37

So I've got something for you here because we wrote this together, yeah, and I did those horrible tables in the S5 2022 guide. So if you look at that table, it says it's after 4 p.m. Yeah, you do exercise, you have three options. You can one reduce your basilintin by 20%, whether it's long night, you know, whether it's yeah, um basilintin. Number two, if you're less than 10, 20 grams of carbs, and number three, very important, especially if you don't like so you call it 20 grams of protein, especially whey protein, yeah. And if people are listening, why is that the case? The theory would be is that the amino acids that get broken down by um the whey protein, specifically, what's the big amino acid? Ah, it's gonna do me nothing. I'll algorithm. No, the in the anabolic one.

SPEAKER_01 54:22

Okay.

SPEAKER_00 54:24

One of the amino acids that should know you're better. Wow, the dietitian, that's shocking for me. Um, but one of the amino acids that's with um uh a signal for M2, a very strong, is also one that goes to the alpha cells in uh Pampreas and supports glucagon release. So just that little minus little bit of glucose, many glucagon goes to that for you can get it from your own Pampreas, presuming you've got enough alpha cell um capabilities, yes. But if people don't want to have cats before bed and they want to have protein, especially within days of GLP1s, for example, 20 grams of protein, way protein shape, and it's not it's not simple building. No, no, no, no, but it's but it's really but that's the deception. The deception is if you take whey protein in a pot of mussel, they thought it was that easy. But that's a really key point. They got half a million data points. They just got our tables from 2020, it's all there. Exactly. I'll put it into our thousand people.

SPEAKER_01 55:16

And and this holds through. We still are recommending in a 2020 position paper this this protein story to overcome or avoid high po, but one recommendation from my perspective is not correct. We have given them this position paper, and so stay on an elevated Q case target. I think if you have this increased risk of high co in the post-exercise period, stay for another two hours on the 150.

SPEAKER_00 55:44

Um and are you saying a bold reduction if the meal is within that nasty period as well, probably? Like 20% off on AI.

SPEAKER_01 55:51

Yeah, so we we are still recommending the 25%. And it's maybe 50%. Like it's just individual response. But at the same time, we just finished the study now where we did this as the pre-pre the pre-exercise roundell dose reduction, would we are recommending 25% reduction? And it might be for one system analog. But I have to say now it does not hold true. Do you need 50 or less? Full bubbles.

SPEAKER_02 56:15

Full bubbles analog, too? Full bubbles.

SPEAKER_00 56:17

Because again, there's where you can put hand it on like 90 minutes before fix it. So I mean, that's not gonna be lighter, does it? That's a problem, besides the development of reality. But anyway, that's cool. So that's interesting. So we basically cover it off there. I think it's also very important. Yeah. When people are vigilant, which you're doing the exercise, they manage to find sleeping clean. Yeah, I didn't do anything like, but then afterwards, actually, that's the key point where your guard is down, yeah, and at three o'clock in the morning, you've now got some strategies. I'll put together a leave piece for for for people so they can know what to do. So I respectfully time, first of all, for the support you give me, the people set of lies for calming me down in random six o'clock emails. I remember giving you a voice value was like, You're gonna have to type that down because I can't understand what you're saying you know too fast. Um, but also just having an interest and being willing to um yeah, just like have a conversation and it's like we just had here, people listening to this. I hope we'll tape away. CGMs are amazing, but still have in your mind, if in doubt, get the get the thing with Brooker out, and especially if you're an exercise lens and on forces in a massive one. We now know that some systems may need to five percent more timing range, although maybe that system is now out of the sense where you don't need that. We can't wait for the data to see that. And we've also now got making sure the post-exercise period is done. I'll give you another reason why I think in the morning um you're not seeing the drop, is because during the day, you've got an accumulation of bolus incidents in long tails. Yeah. So you're not you're not actually a true basal condition at any point. If you say for 0.2 grids per kilo of bullets, it's a six-seven-hour tail. Yeah. If you have one at 2 pm and then 6 pm, yeah. Like that's going to midnight and it's not basal conditions. So I'm not incident wars, my figure than that. So I'll leave that there. But I just want to say thank you, ATCD. Any anything major that you've picked up for ATCD that you're excited about? Obviously.

SPEAKER_01 58:13

I'm not allowed to say I'm not allowed to say too much, but we were breaking on the the artificial banquet consensus. Okay. Yeah.

SPEAKER_00 58:22

Or is that what we'd all like the weird different, potentially different, you know, what the definitions are.

SPEAKER_01 58:27

But at the same time, we saw some data published of fully telescoped. So this was quite of interest. And we saw some great presentations around exercise from Mike, Desi, Klemmer. Um, so from my perspective, I think the future is promising with all the technology, with novel insulants maybe coming up next. And at the same time, I think the most important part, do exercise.

SPEAKER_00 58:47

Yes, exactly. From the first from the first thing with anything, I've had loads of videos that have the most watched thing. People came and said, I can't believe someone said they would rather have a HVMC of 7% and be really active versus a HVMC of you know, um 6.5%. But but the crumble, yeah, exactly. All right, so thank you. Thank you for listening. And what I've learned from ATTD is how to optimize some of the AID settings for uh tandem, all of them, and we're gonna then put that in the AID series turn up tune. So thank you for cheering in. Done.