Despite of the press hype, the Australian children were just being fitted with a MINIMED® 640G plus PR hype from an oz researcher involved in the algorithm. I would believe that this could possibly be helpful in preventing night time lows. I am on MDI, but my basal is 12u Lantus or roughly .5u / hr and 1u of insulin reduces my bg by roughly 1 mmol/l. Remember also fast acting insulin from a pump or injection has its effect roughly over the next 4 hours
So if I shut off the pump basal when bg = 5 my insulin on board will make me bottom out at 3. Well its better than going down further. But if I walk for an hour with only my basal active I would expect my bg to drop by 8. So even if my pump is shut off at the start of my walk, my insulin on board will still lead to very low bg.
So I do not expect a pump that just reduces/stops insulin to cope with exercise. This means looking ate the dual systems which also pump glucagon. Currently glucagon is unstable and a fresh solution needs to be made up each day. It is also expensive.
exactly. Manually shutting off the pump when I’m on a hike doesn’t always stop me from going low enough that I have to treat a low at the midpoint of the hike. It “helps” but doesn’t prevent lows needing treatment. Furthermore, if the sensor is reading lower than it should (a COMMON problem when sleeping on the Enlite–pressure on the surrounding area often causes interstitial fluid to move away from the sensor), then anyone with a relatively high basal rate at night, will likely end up with a HIGH bg by morning thanks to a mistake made by the pump/sensor. I would never personally enable the SUSPEND ON LOW feature for myself, given how I sometimes get false-low readings at night. Now that I tuck my arm (with sensor located there) under my pillow, I’m not as likely to get a false-low as if the sensor is pressed upon directly by my head, but even so, laying down, in and of itself, seems to occasionally lead to false lows. I HATE waking up higher than normal and most of the time I don’t.
You can’t generalize the way you appear to be doing. I am willing to believe whatever you want to tell me about your body because, well it’s your body, your business. But there is so much diversity of experience in diabetes that what you experience is pretty much meaningful mostly for you. Others could turn off their basal for a low glucose result and even with vigorous exercise still see their glucose levels rise.
Yes. the 640G is not an “artificial pancreas”. But while it’s only an albeit simple step in a rather obvious direction, I’m still intrigued and hopeful about it. I’d be very interested in trying one out. Whether I do or not depends on what other options are available in about a year & 1/2 when I may get to choose. But it’s certainly at the top of my list at this moment in time.
That’s a pretty strong assertion considering you haven’t yet had any experience yet with the actual technology. While it’s true the tech is similar to what you now use, it is not the same. Medtronic has tweaked it. Possibly they have even improved it. Who knows? It could happen.
I am saying that your pump is not a 640G. Or more precisely it’s not whatever tech Medtronic might be selling a year (or so) from now. So saying never strikes me as a little premature. That’s all.
A good point that people are different. Sometimes with my usual non vigorous exercise my glucose shoots up, but it usually does not, so I can not rely on this effect. Some people seem to think that more vigorous exercise is more likely to provoke a glucose spike. In that case the 640 g will notice the spike and up the insulin to correct. However some people report that some time after vigorous exercise the bg drops quickly as the body recharges. This will also be a bit of a challenge for the 640g as it has no way of upping the bg…
I am not a cgm user, but I am encouraged by reports they seem to be getting better. But the proof of the pudding will be the using. Any 640g users able to tell us how it goes with exercise?
If you are in the USA you are presumably using a 530G which is essentiall the same as the old Veo. The low suspend feature on that was pretty rudimentary and unreliable, more an emergency feature to try to prevent severe nighttime hypos. Most of the users I know used to turn the feature off. The 640G is quite different. The low suspend is predictive, and basal delivery is restored on the basis of sensor trends. I know one user. She was quite sceptical but has found it very effective in practice. She sees several short-period suspends most days. She also hasn’t had a significant hypo in more than a month.
What part of “I don’t use the feature because the Enlites can give false lows, especially when lying down” do neither of you get? I do not trust CGM technology enough to have it cause a pump to suspend delivery during the period when I’m most susceptible to getting high bg’s. There is no algorithm that is going to make up for flaky sensor technology. Once the sensors are made an order of magnitude more reliable, then we can talk.
Insulin lasts for 4-5 hours (roughly) so suspending when low is a bit of “the horse already left the barn”. If I have bolused even a little bit too much, the extra IOB that’s active could, and has on countless occasions , caused me to go low. Suspending (assuming ACCURATE sensor data) is not going to prevent the low.
Your anecdotal comments hold no sway with my decision to not use the suspend-when-low feature on my current pump or the 640G.
We’re all aware of the time lag when delivering insulin. I don’t use Medtronic pumps anymore but I do have a basic understanding of the “suspend when low” model versus the “suspend when predicted low” model.
@phoenixbound, I agree with your lack of trust in the suspend when low model, especially given anomalous compression outlier lows when sleeping. I suppose when the system works as designed it can prevent a several hour low that could be catastrophic. In the past I have slept through lows enduring 3-5 hours. I shudder to think of that now.
The predictive low suspension, however is another animal, one I think should perform much better. Especially if your variability is lower. If, for example, that a suspend when a low is predicted pump sees when I’m at say 90 that it expects me to cross 70 in 90 minutes given the current slowly sinking trend and maybe some leftover IOB from evening snacking. In this case, if basal insulin is suspended 90 minutes before I hit 70 then I can see this as a timely and valuable tactic.
It would even be better if the algorithm can predict the ultimate timing of the BG upturn and turn the basal back on in time to counteract the predicted rise. This is something I do manually and perhaps this is where the human brain can still outsmart the current state of the art in automatic BG control. It’s thinking two moves ahead of the trend. I don’t see any reason why a predictive system like this could not learn my personal BG quirks and become smarter over time.
I think the key to making the current predictive model work well is keeping BGs less variable. Shallower slopes gives you the luxury of time such that pump basal shut off can actually protect a low-BG threshold. In other words maintaining a gentler slope on any downward trends. I don’t think this is too difficult to set up. Since I limit carbs, I almost always maintain low slope BG lines overnight.
I am very interested in how these early iterations of the basic artificial pancreas work. These are baby steps but must be taken before more sophisticated and reliable technology arrives. For me I came to the surprising conclusion only just recently that I will probably not participate in any of these early systems since my manual system will probably not be technically surpassed for several years to come. I am confident, however, that this technology will evolve quickly and a very good every day system that will easily exceed my human mediated methods is just around the corner.
It is a device just like pumps and CGM most people will not be able to get because insurance will not pay for it and the cost of supplies( I have good insurance and sure I can afford one for my daughter). It will no doubt have some of the same challenges as current pumps, site insertion and insulin absorption problems. Who wants to have two three or more devices attached to them? I know my Brother and Daughter both do not want devices attached to them. My daughter will wear a CGM though. I also find it interesting people will not use their CGM for blood glucose readings and dose (Dexcom G4 is using the same algorithm as the AP) But people will trust the AP and sensor combination to be able to adjust the glucose levels correctly. Personally from my perspective Smart Insulin and even companies like Viacyte and Sernova stemcells options will be much better than an AP and could be here as early as 2020.
I hope you are right. Having said that, I looked at the Smart Insulin patch and must say it does NOT look “painless” as they describe. It looks downright irritating/painful. Even the failed Glucowatch irritated its users and didn’t look anything as intimidating as Smart Insulin. If it works well, I hope I’m wrong about skin irritation/pain.
depends on the level of irritation. I don’t feel my sensors and I don’t feel my sets. If I did, I’d not be so willing to wear either one. I have enough irritations in life, having PN–I don’t want to add another.
Measuring glucose? I expect that like current CGM’s, it’s going to measure interstitial fluid instead of capillary, so there’s the first problem–a delay in glucose response, compared to blood samples. Secondly, NO freaking sensor yet on the market is accurate enough to be trusting our lives with it. Even if they came with an “odd-man out” method (think ‘Minority Report’) of determining glucose level, I’m still not on board. Cost, with glucagon included? Prohibitive.
They need to go back to the drawing board. The FIRST thing they need is a SUPER-FAST insulin with just minutes of duration. Then they need redundancies to put the Space Shuttle to shame. Then they need sensors that can detect BLOOD glucose levels instead of the delayed glucose present in interstitial fluid and be at least an order of magnitude more reliable than CGM sensors.
And by cgm you mean Enlite? Okay. Nobody would dispute that cgm ( meaning Enlite ) is not good enough. But if there was a device that could use interstitial fluid and present blood glucose levels accurately then that information could be used to adjust insulin dosage. The goal would then be to keep blood glucose in a relatively normal range. Medtronic seems to be bungling it. Their devices are called artificial pancreases. But I was disappointed after waiting for their second generation improvement. It was more like than different from the first generation.
