Pancreas broken? Make your own.
July 20, 2017 9:28 AM   Subscribe

Dana Lewis has Type-1 diabetes so, she went ahead and built something for that. She built and programmed a computer to monitor her blood sugar levels and adjust her insulin levels automatically. Then she founded Open APS to be a repository for her design, and to let other people use it. She has gotten plaudits from both the maker and the medical community. On her twitter account, Lewis often uses the hashtag #wearenotwaiting. Is she the first wave of a breaking trend?
posted by lumpenprole (12 comments total) 29 users marked this as a favorite
 
I have a friend with type 1 diabetes and an insulin pump. I'm kind of shocked that the glucose sensors exist, and the pumps exist, but no one (I don't think) integrates the two with feedback control. I figure it's a patent or liability issue.

OTOH, I'm sure there were lots of liability issues with the first insulin pumps, and suppliers got over them and they're making a sweet buck now. So lets go future Dean Kamens, the market calls!
posted by Popular Ethics at 9:48 AM on July 20 [3 favorites]


There is at least one FDA-approved CGM system already on the market, I assume there are others as well. Generally speaking the hangup on getting CGMs widely adopted is the necessity of invasive testing - a wearer must have two stents, one for the monitor, another for the pump.

Years of work on non-invasive monitors using various metrics derived from attempting to analyze information from the wearer's skin (glucose content of perspiration, changes in various aspects of the chemical content and refractive density of the epidermis, stuff like that) have produced meters that kinda sorta work.

My wife uses the pump half of the Medtronic system I linked to, but does not want to try the CGM, in part because she believes that her fifty years of experience monitoring and testing her blood glucose is a more reliable system than the Medtronic CGM, and because she really is resistant to the idea of *another* needle hole to monitor for infection and blockage. Her skepticism also reflects years of dissappointment with regard to the development and usefulness of noninvasive CGMs.

In my understanding, the end-run that Lewis is attempting is not so much around an impossible technical blockage preventing the technology being made available to end users as it is around the expense associated with it, the clumsy and inflexible user interfaces associated with the outcome of medical-use testing approval, and the not-incidental vertically-integrated consumables and device lock-in strategies associated with the big medical technology companies that have the resources to meet the requirements for device approval.
posted by mwhybark at 10:10 AM on July 20 [9 favorites]


My wife is a Type I that has been using the Medtronic pump since the 90s (when it was called Minimed). She hated the first generation of CGMs. She is due for a new pump this fall and was telling me CGM has gotten better. I wonder if she would let me build that artificial pancreas piece for her?
posted by COD at 10:11 AM on July 20


Ah, I see, she is bridging the gap between the CGM and the pump such that the CGM triggers pump adjustments. When we last looked at the system I misunderstood it to offer that feature, unless it did and then the feature was pulled (unlikely, I think).
posted by mwhybark at 10:14 AM on July 20 [1 favorite]


There are several CGM systems on the market. It's a very well-established technology at this point, and virtually every T1 diabetic know uses one. From what I've seen (my spouse is diabetic and very involved in the local community), the Dexcom is much better than the Medtronic. So if someone has a bad impression of CGM based only on experience with Medtronic, I'd urge them to look at Dexcom and maybe the Abbot Libre (which has some limitations, but also some benefits).

But this discussion is beyond CGM, and into the combination of CGM+insulin pump. AFAIK Medtronic is the only company that currently has a commercial FDA approved closed loop pump/cgm system (the Minimed 670G). It just came out, so feedback is still early, but I know several people who have used Medtronic systems before and are avoiding the 670G because of their previous experiences. However, there are other companies working on devices like this, so I think we'll see a lot more non-DIY options in this space soon.

I do know a few people who use OpenAPS and they seem happy with it, but it is definitely not for everyone. I do think it's awesome that people have found ways to DIY this, when the medical device companies are moving too slowly.
posted by primethyme at 11:38 AM on July 20


One other thing to point out. To have a true "bionic pancreas," you need to be able to deliver both insulin and glucagon. The current systems I know about can only deliver insulin. Beta Bionics is working on a device that can do both, but I think it's going to be a while before it's available.
posted by primethyme at 11:40 AM on July 20 [3 favorites]


This guy (I swear I read about him here on blue) talks about why you'll pretty much never have a non-invasive BG monitoring device. (Author passed away this year).

And put me down as someone who would only try the medtronic CGM if it were free for the first few months.
posted by k5.user at 12:00 PM on July 20


mwhybark: "she really is resistant to the idea of *another* needle hole to monitor for infection and blockage."

Is there some reason both monitoring and injection couldn't be done through a single hole?
posted by Mitheral at 6:58 PM on July 20


Is there some reason both monitoring and injection couldn't be done through a single hole?

Cross-contamination. If the needle has just been used to inject insulin and then aspirates "blood" to measure glucose then what it will actually aspirate the fluid it has just injected, giving a false reading for the glucose.
posted by *becca* at 3:05 AM on July 21 [1 favorite]


I've used both Minimed Enlite and Dexcom CGMs, and Dexcom is vastly superior in my opinion. It's been a life-changer. For anyone thinking of trying out a CGM, don't give up if you've had a bad Enlite experience.
posted by chinston at 10:38 AM on July 21


Okay, so I am not a fan of almost any medical device company but reviewing some of the worst case outcomes from device trials gone bad that came up during my CITI training made me appreciate how hard it is to sufficiently test a device in a way that adheres to modern standards of medical ethics. There is a lot of potential for harm when it comes to closed loop insulin regulation, and a lot of opportunity for inequitable trial design.

That being said, the diy'er in me loves this. Also I hope artificial pancreas are a thing tomorrow.
posted by midmarch snowman at 7:58 PM on July 21


Scott Hanselman talked to Dana on his podcast last year and has posted a few times on his tech-focused blog about the state of diabetes technology (Scott is a Type 1 diabetic).
posted by markr at 1:23 AM on July 22


« Older The Medieval Spanish Chef   |   US senators seek to make boycott of Israel a... Newer »


This thread has been archived and is closed to new comments