Just had my Omnipod 5 training. I’m intrigued that the system changes your underlying basal rates with each Pod change based on your previous pod sessions. My question is: is there any way (perhaps using Glooko or the PDM) to see what the changes it makes are?
Also, does anyone have a sense of what it’s changing and how? I’m referring to the pod to pod changes not the 5-minute adjustments that are based on your predicted BG. For example, is it simply upping your underlying basal from the previous sessions if it sees that you were consistently needing extra insulin to correct for highs or does it adjust other things like insulin sensitivity if it sees that the corrections it made consistently didn’t produce the expected result?
I had never heard this before. I will be interested to see what the powers that be say. I am finally off U500, after too many trips to the ER. My problem now is that I am changing pods almost daily. My doc says she has people that don’t use pod all the time to stretch out the time the pod stays. Any other ideas?
I have some ideas and questions for you, @Dennis4. I suggest that you post your discussion as a new topic instead of a comment here. This thread is better suited to the original poster’s question. If you post a new topic, I’ll respond there!
I’ve been on Omnipod 5 for about a year and a half. I would not consider myself a true expert, but I have really tried to learn about the pump to get the most out of it. Here is what I would say based on my understanding of the system: the algorithm depends only on the total daily insulin you use - not insulin-to-carb ratio or correction factor, etc. I know you input an initial basal program, but actually the algorithm completely ignores this once you’re up and running (I think after the first pod change), unless you put the pod in manual mode or the connection to your Dexcom is broken for a long time. So when you ask “can you see the changes it has made?” I assume you mean to parameters like insulin-to-carb ratio, and those parameters just aren’t used by the algorithm. The algorithm itself is very opaque, and I think the fact that it does not care what basal settings you input is a bit of a downside for those of us who would like to optimize (for example, before exercise).
Also seems like you know this, but the system only changes its algorithm settings when you do a pod change AND have worn the previous pod for at least 48 hours. And the whole thing is tied to your controller, so if you lose the controller or need to change it for some reason, you start from scratch.
The one output you can see is how much insulin the algorithm has delivered, and the choices it has made every 5 mins based on your Dexcom reading. It’s under History Detail - under Summary, you can see the total insulin for the day, and the percent of that insulin that is basal vs bolus. Under Auto Events, you can see your Dexcom reading and the insulin delivered by the algorithm every 5 mins. Interesting, but not the same as a summary insulin-to-carb ratio.
And finally - overall, I do really like the algorithm. It has taken a lot of the mental load off of me and has massively improved my sleep. But I HATED it when I first started on Omnipod 5. It took a while for the algorithm to “learn” my insulin needs and get good. A few weeks at least. I think I could have had a much better experience if I had listened to this podcast before starting on Omnipod 5 or if my endo had known more about the system to help me. Point is, listen to the episodes before starting if you can and have a little patience as you get going! Hope you love it.
Great advice. Thank you!
I did happen to listen to one of the episodes of the Juicebox podcasts on the Omnipod a couple of days ago. (The one about the algorithm where he talks to the woman who writes Omnipod explanatory materials for Insulet.) Unfortunately, this was after I’d had my initial training and had input all my basals etc.
But, whatever I input was ok, because it’s actually worked ok for the first couple pods.
The algorithm is indeed quite different from Tandem’s. I feel like it’s sort of a middle ground between Tandem where the user controls a lot and the algorithm sort of tweaks it moment to moment without ever altering the structure and the iLet which sounds like the user has almost no control and the algorithm over time figures things out.
I don’t love that Omnipod doesn’t really share what it’s doing on a meta level; I look at the “auto events” constantly and can see what it’s doing moment to moment which is often a surprise. I feel like I’ve already learned from it. I kind of knew that I was feeding my basal when I was on T-slim but the Omnipod has already shown me that I can pull back on my basal significantly.
A change is good, and, even if I ultimately go back to Tandem, I can bring what I’ve learned with me.
