So, since Samson was diagnosed he’s maintained pretty good control, thanks to the artificial pancreas algorithm he uses (openAPS), but there were at least two or three times a day when we had baked in lows we’d have to preempt, where the IOB was listed as negative and yet he still kept dropping precipitously.
I tried tweaking so many things but really, nothing worked for long.
About two weeks ago I did something that others had been mentioning for a while. I thought I had tried these settings and found them not to work, but I had tried everything else and just figured, WTH. So, I changed his DIA to 6 hours up from 5 hours. Word on the street is that this is the default DIA set for the Medtronic 670G’s closed loop pump, and I figured that it was worth a try.
I didn’t think it would make such a big difference; after all, 5 hours is already much longer than many peoples’ programmed DIA. We’d tried a 5 hour DIA when he was first diagnosed and it was clearly too long.
But the results have been really amazing. No more baked in lows. He just drifts down very slowly to target range over the course of the day. And numbers that used to have me reaching for the glucose tabs now mysteriously just settle out right at 80. For the past week or so he has only gotten a glucose tab at daycare every 3 days or so, which is a huge improvement over once or twice per day.
Practically speaking, his average numbers haven’t changed much (he’s up a few points from his low of 122 mg/DL), but he is having a lot fewer lows during the day, with less than 4% time spent below 70 now. And the quality of life change is HUGE. The level of stress we feel has gone down tremendously. Not feeling like I have to constantly think about the potential need to feed him has really changed our family dynamic and he’s been a lot cheerier too.
Of course, now he’s actually settling in at his target BG, rather than crashing down towards 80 and being corrected up to 100. That means he’s spending more of the day sitting at 120, which is raising his average. So once I decide this is a stable change I may lower his target. And he’s having more postprandial spikes up to, say 155 or 160 rather than 130, which suggests I may need to strengthen his carbFs after having changed his DIA. But it feels great to know that I can set a target and he has a reasonable chance of converging upon it.
We still haven’t gotten perfect night settings; he’s still drifting down towards 80 and sitting there for hours every night despite his target being 120 at night and having many hours of zero temping. (80 is a great overnight BG of course, but if your target is 120 and he’s not converging upon it, that’s a sign the settings aren’t tuned properly). But I feel like we’ve definitely had a breakthrough.
Anyways, just wanted to share this because it’s the kind of thing I found really simple but also transformative. My theory is that if the DIA is too short, with a normal pump you may get some insulin stacking but it’s not that pronounced. But with a closed loop algorithm that is constantly piling on high temp basals and then zero temping to compensate, this little discrepancy between the programmed and actual DIA adds up to big differences in the last 4 to 6 hours after insulin was last administered.
@Michel, I think you talked about kaizen, or the notion of continual improvement. I think that attitude is what allowed me to get to this point without just settling for a status quo. Of course, sometimes you also need a little breakthrough, and changing the DIA was a bit of a ‘click’ moment for us.
Anyways, just wanted to share.
. Anyone else?
I’m sorry, that’s a tough one.