Frustrated with current pump options

Unlike the 670, if control-iq is turned off you lose the low glucose suspend feature. At least with the 670, you’re still protected with the low BG suspend when auto mode is turned off. This is why I suggest using basal-iq for a while to see how their BG responds. If they find the loop function doesn’t work for them, and turn it off, all ability to have a low BG suspend with control-iq is forfeited and there is no way to get basal-iq back into the pump.

1 Like

The sleep function does have a tighter control threshold but remember, when this feature is active there are no correction boluses given by the pump. It only adjusts the basal rates. I find it very strange that Tandem has not said one word about how much the algorithm will raise or lower the basal rates depending on the predicted BG. The only thing they’ve said is try it out and see how it works, and control-iq is not going to work for everybody. These are not the responses I would expect. Losing basal-iq to get something that has not been explained at all makes me very nervous. The trials results data they keep throwing out is pathetic at best. Seriously, the group with control-iq were in the range of 80-180 71% of the time, compared with 50% time in range with the group without control-iq. Those are not results I would be celebrating.

I spent nearly an hour and a half speaking to Tandem tech support today asking various questions about Control IQ. Here’s a pretty significant take away:

In the X2 simulator goto My Pump -->Control IQ–> Control IQ. There is as setting for “Total Daily Insulin”. On the simulator this value maxes out at 100 units. I had tech support that on a real live pump that it also maxes out at 100 units. This value is important…read on…

Refer to page 261 in the manual (Page 263 in the PDF version) --> Maximum Insulin Delivery:

"When Control-IQ technology predicts that your glucose value will be above the upper limit of the target range 30 minutes in the future, but the maximum rate of insulin delivery has been reached, Control-IQ technology stops increasing the insulin delivery rate."

Once you hit a system maximum of 100 units of infused insulin in a day (or whatever you as a patient and your doctor program as the maximum per day), Control IQ will only provide basal insulin reductions & suspend basal it will not increase your basals above your personal profiles, nor presumably provide correction boluses. I confirmed that with tech support. I read that as a kill switch to shut down the A.I. algorithm to protect the user from being over-medicated.

My take away is that this was built in to the pump to shut down the algorithm from potential over delivery in case your CGM was incorrect or from insulin stacking which is a good protection, however if you average near or above 100 units a day of insulin, or just have a bad day in general in terms of diet or illness, and near that set ceiling, Control IQ will not do anything to reduce high sugars, the patient will have to manually bolus.

The live event today was a sh!t show. They didn’t answer any questions with specifics and just read the bullet points from the website. The video was terrible, it kept freezing and cutting out the volume. They were supposed to do a live question and answer afterwards but the video froze and it just ended. I’m done with Tandem, I’m eligible for a new pump in March, I’ll be moving to the 780 which will allow me much greater control over the functions of the pump instead of shoehorning me into a cookie cutter setting.

What is known about the 780G at this point other than it being Bluetooth enabled?

They’ve improved upon much of the features on the 670. The big one will be that it’s much easier to stay in auto mode. The target BG can be customized as low as 100, instead of the fixed 130. There’s a lot of nice changes that have been made. This is what the 670 should have been right out of the box.

Re: Total Daily Insulin limit in the Control IQ menu

I reached out to Tandem customer support to try to verify whether the 100 unit limit referenced above is the incremental total insulin that Control IQ will administer above and beyond profile basal & manual boluses or is it universal (total daily insulin from the pump regardless of manual vs Control IQ administered). Should have a confirmed answer in a few days.

When they call you back, ask about the weight parameter too. There’s no reason for this to be in there. The pump is using my ISF, my basal rates, completely ignoring my personal IOB timeframe setting and locking it into a mandatory 5 hour slot. These parameters are all the pump needs to regulate the BG. What does weight have to do with it?

Jason,

The pump has minimum requirements to enable the Control IQ feature, no children under the age of 6, weight of 55lbs, and total daily insulin 10 units; that appears to be a safety feature.

Per the product manual:

In addition to the required Personal
Profile settings, there are two values
specific to Control-IQ technology that
must be set. These are:
• Weight
• Total Daily Insulin

An additional inquiry that I’ve made is for more information as to how the Control IQ automated basal adjustment feature works, We know it will adjust basal rates up or down if a high or low glucose reading is predicted. My question which is not addressed in the Tandem manual or marketing materials is how aggressive are the basal adjustments. For example if my basal rate is 2 units per hour, and a high bg is predicted and/or sustained, what is the upper range of basal adjustments built into the system, i.e. where does it max out?

Is it limited to increasing basal rates by 10%? 25%? 50%? 100%? 200%? etc. If other folks are considering Tandem due to Control IQ it would be helpful to the community if you press for these answers so Tandem knows we care about the implementation.

From what I recall reading from the DIY LoopDocs, that system uses temp basals that may be many multiples of your normal basal in order to control predicted highs and/or to bolus for meals. I’d like to know how Control IQ compares.

I’m not sure if I’m crazy to even be considering this but I keep wondering whether the DIY Looping via LoopDocs is a better mousetrap. Is it worth experimenting and taking the (significant!) risks associated with it and also dealing with the subsequent back end management of the application.

I like the Paradigm pump as a pumping platform and part of me is very hesitant to switch without knowing all the minutia as to how Control IQ will work. I’m not thrilled with a change in reservoir / cartridge filling procedures, as well as dealing with another device to charge; but if Control IQ delivers the goods it would be worth it. On the other hand, Looping will require a Riley Link module that will need to be on my person at all times and also charged nightly, as well as an obviously working iPhone at all times. That creates 3 additional points of potential failure in that system: the app, the Riley Link, and the iPhone. That is in addition to the pump and CGM sensor were I to proceed with the Tandem T:Slim X2. For anyone who does any serious degree of traveling it is a consideration.

My head is swirling right now.

@trifona, it sounds like you have put a bit of effort into this decision, and are working various solutions around the currently available options (yup, I am stating the obvious). Sometimes during a design, it pays to take a step back and review what you believe to be “critical to success” . I am not sure any feedback pumps are going to get you below 7. The simple reason is the pumps will be set conservatively because a severe hypo on a feedback pump leaves a company open to litigation…(period) and that’s not going to change… ever, And I am guessing that it will never be as aggressive , even with an AP pump with that new glucagon (giving the pump the ability to raise blood sugar).

So if your most critical goal is be more aggressive overnight without killing yourself, IMO, focus on that. That can be accomplished with Basal IQ on the Tandem, or with the equivalent “Guardian” on the Medtronic, or with a DEX on any pump and a predictive low alarm.

If your most critical goal is for AP (artificial pancreas) automation and for aggressive control, IMO it will not happen using conventional means for many years and hinges on much faster insulin with little or zero tail, and an order of magnitude increase in accuracy, response speed, and reliability of the CGM, yes including G6… maybe 2 orders of magnitude for the Guardian. This may not be in our lifetime. Well ok, maybe in yours… I am pretty old.

I hope you are well and I wish you good luck.

1 Like

Joe, thank you for your well thought comments, especially those regarding “critical to success”.

So if your most critical goal is be more aggressive overnight without killing yourself, IMO, focus on that. That can be accomplished with Basal IQ on the Tandem, or with the equivalent “Guardian” on the Medtronic, or with a DEX on any pump and a predictive low alarm.

I love the simplicity of the compact but robust build quality of the Paradigm pump. I love the tactile buttons, the intuitive menu structure, the ease of cartridge changes, and the availability of AAA battery power that can keep me going for 2-3 weeks. I love the Dexcom G6. I wish they spoke to each other, but they don’t without DIY Looping.

I don’t have a lot of trust or interest in the Guardian sensors, they seem to be a lot of hassle to me which is why I have been dismissive of the 670G for the most part.

@trifona You and I are definitely on the same in regards to Tandem and the lack of information being being given about control-iq. I have been very curious about how much the algorithm raises and lowers basals. I’ve been asking both tech support and marketing to answer this very simple question. As of yesterday, after that clown show they put on, Tandem is still refusing to answer the question. That live stream yesterday was supposed to give us the answers we’re looking for. Instead, we got a horrible video feed, a marketing manager that did nothing but mumble, a slideshow that was so blurry you couldn’t read it, and a question and answer portion that never took place. I believe this was done on purpose because most of the questions being asked were the ones Tandem seems to have no interest in answering. I think it is very reasonable of us to inquire about how much insulin this algorithm will be giving us. Like you said, a ballpark number of 10, 20 percent would be an adequate answer.

The problem with DIY looping is the pump itself. Getting your hands on an older Medtronic pump with a very old software version will be difficult. My MM 723 I got in 2010 is running a newer software version that you can’t use for DIY. So, you’ll have to use a very old Medtronic pump that has most likely seen many years of service, and reliability becomes an issue. Especially, for something you’re going to have to pay a good chunk of change for with no warranty and no way to repair it.

@Jason1 Unless I already have good condition pump in my closet…In that scenario I just have the cost of a Riley Link and Apple developer license to deal with.

:grin:

@trifona If you have a Medtronic pump that will allow you to use the DIY loop, you should try it. The low glocuse suspend in the DIY loop is based off of Tandem’s basal-iq. You can taylor all the settings to fit you and not be stuck with features you can’t customize.

@Jason1

“I have been very curious about how much the algorithm raises and lowers basals. I’ve been asking both tech support and marketing to answer this very simple question.”

Some answers. Tandem Sales called me back this evening and indicated that there is no hard ceiling maximum other than the pump’s ability to deliver 15 units per hour. 15 units per hour via basals that would adjust every 5 mins based on predictive values. The actual ceiling would be determined by a combination of personal profiles, Insulin On Board, sensitivity factors, etc.

I’m somewhat confused / incredulous that the system would be that aggressive when Control IQ’s bolus corrections are limited to only 60% of normal correction factors. The other reason I’m confused by the information is this snippet from the manual on page 270 (re: Sleep Activity):

“if you sleep less than five hours, Control-IQ technology may not be able reach or maintain CGM values within the sleep target range.”

If Control IQ is capable of super aggressive basals of up to 15 units per hour why would it also need at least 5 hours to get the patient in range? Your glucose would need to be so incredibly high (G6 maxes out at 400) and your sensitivity would need to be so very low unless I’m missing something.

1 Like

@trifona Im more confused now than I was earlier. I can’t imagine the algorithm being that aggressive. The lady in the live stream yesterday, stressed that you still have to do correction boluses and that the system will only try to keep the BG from going to high. The 60% bolus once an hour is not aggressive at all. The pump doesn’t do anything until the BG is 180, then it only gives a bolus with 40% of the insulin missing. You’d have to constantly bolus on top of a bolus. This algorithm is definitely designed for those who can’t get their diabetes under control. You can’t fine tune any of the settings to keep the BG at a 100-120 range by itself. Why they lowered the low BG suspend from 80 to 70 is beyond me. This is what happens when you’re using data and advice from the antiquated and useless ADA, who still think 180 is not a high BG.

This is so frustrating that Tandem has botched this rollout so badly. They withheld the emails going out an extra week so they could do this live stream yesterday and inform their patients of how control-iq works. They didn’t explain anything that wasn’t already known. Supposedly the emails to start the update process will begin next week, but will be staggered over the next few weeks and months. Not being able to go back to basal-iq once you update is causing a lot of people to rethink control-iq.

@Jason1 I’m confused too, and I also suspect that he algorithm is not that aggressive.

I’m less of mind that they “botched” it and more thinking that these are very hard topics and concepts to communicate to the general public. When was the last time a company had to advertise an algorithm update? I can see the POV that they keep things simple on the website, marketing materials, etc because in essence they are advertising the benefits that their algorithm can provide to the patient. However I do wish the official product manual had more detailed specifications around how this algorithm actually works.

When the 670G launched Medtronic didn’t communicate squat about their system…it was “Auto Mode” magic beans; the patient disappointment came later. The MDT algorithm appears to be extremely conservative, the dummy protections high, and the sensors are poor.

Control IQ is next up to bat and I don’t doubt it is better than 670G’s “Auto Mode”, thanks to having time for additional development, and to drive enhanced acceptance from the FDA for letting go of some of those dummy protections. Thankfully Tandem has partnered with Dexcom so you know the CGM is best in class. Do i wish they communicated more? Yes for sure. I think what they have communicated to date is probably enough for 99% of those out there. The problem is folks like you and me are being tempted by what we see and read about with DIY loopers and have higher expectations. Hopefully they learn from this and improve but I find it hard to criticize them too much in how they are marketing an algorithm.

I’m also realizing that I’m subconsciously so mentally attached to my current insulin pump regimen that I keep looking for reasons to shoot down alternatives.

I agree with you, but we’re at a point where we need more info other than, the pump will increase your basal rate. It is an algorithm and not the next new car or gaming console, but they are showboating this update and just need to get it released and put out more comprehensive information regarding what the algorithm will be doing.