Are the DIY Artificial Pancreas Folks Outlaws?

I just read this piece from Dana Lewis’ blog.

Dana, if you don’t know, is part of the team that has put together a closed loop artificial pancreas system via “hacking.” They call it a DIYPS (Do it yourself pancreas system) and it’s not currently following any FDA approval or guidance. It’s sort of renegade.

You can read more on their website. I was particularly interested in the TuDiabetes community’s opinion on the blog post I linked to where Dana writes about the attitudes of some folks at the FDA that their hacking approach is dangerous. I’ve honestly had zero opinion on this until reading her piece. The part that stands out the most to me is this quote:

Safety is what drives the #wearenotwaiting movement. In my case, I refuse to sleep another night with the fear that I won’t wake up in the morning because there’s not an FDA-approved system on the market that will wake me up if my life is in danger

So I’m curious what your opinions are on these “renegade” hacked devices are. Dangerous? More dangerous than not having access to them at all?


Are they dangerous? I suspect the danger level corresponds to the DIYer’s skill level. Were I to attempt this it would be a truly dangerous thing. Considering Dana’s and her groups personal stake in this I would think their risks are minimal. Every drug and device comes with risks or side effects. The question is do the benefits out weigh the risks. It may be worth the risk for Dana but for the average Joe it probably would not.

I applaud what Dana and her group are doing. They have a personal motivation for doing this. They say that necessity is the mother of invention. This is how real innovation happens. I do hope that one day they begin the FDA process so that if they are successful everyone can benefit.

1 Like

as soon as I read this: “better real-time BG readings using raw unfiltered sensor values” I knew I was reading something written by someone who doesn’t know the technology. Sensors don’t read blood glucose–just interstitial fluid. Words matter and when they are incorrect in makes me wonder about the knowledge of the writer.

I think that Dana Lewis and others in the WeAreNotWaiting DIYPS movement productively irritate the regulatory/medical tech industry. They’ve never been prodded from this quarter before. The game is changed and the incumbents don’t like it. That combine should feel a bit irritated. They have not moved with alacrity when viewed from this T1D’s perspective. Do you think Dexcom’s recent leap into a higher gear happened in a vacuum? Hooray for the renegades!

1 Like

I don’t think any of these devices automatically dose insulin, so to me what risk there may be would be low.

Also we already accept a lot of risk in standard prescribed treatments (inaccurate meters) that the FDA approves.

It’s my understanding that Dana Lewis uses a closed loop (N=1) system, one that changes insulin delivery based on CGM sensor readings. I haven’t checked in on their website in a while but during my last check her pump set temp basal rates to alter insulin delivery up or down during the night to keep her BGs in range. It’s been a few months since I last peeked; I suspect their project is more advanced now. There’s a whole community involved. It’s organic.

The system is not easy to setup & that for for a good reason. It forces the person building the system to understand how it works & it’s limitations. Would I use a DIYPS, yes! Could I build one? Nope. But what I can do is read through their blog & get a better understanding of some of the techniques used with the system. Mostly ‘eating soon’ & figuring out my carb absorption rate so I can better dose for meals. My A1c is down, so I owe them thanks for sharing that type of info.

And just to note, DIYPS/OPENAPS hasn’t always been the only renegade system out there. Ever read the story of how Bigfoot Biomedical got their name? Bigfoot Biomedical Sets its Sights on Simplifying Type 1 Diabetes Management

I think it is great that they are doing this work however I’m sceptical about creating a reliable closed loop system using the current cgm technology.

For me personally the sensors aren’t very accurate for the first 24 hours and last day of use. I don’t think I could trust them to automatically give me insulin.


Interesting, I didn’t know that.

So I read their summary of the closed-loop version. It kind of falls between having a regular pump/CGM set up and the full-blown artificial pancreas systems the big players are developing. The latter involved developing special faster-acting insulins and glucagons that can last more than a few hours in a reservoir, plus algorithms to raise as well as lower BGs etc etc. Whereas the DIY is all off-the-shelf parts and standard insulin, and doesn’t have a glucagon component. The “closed loop” version still relies on alerts and manual intervention to handle sudden lows and so on. “Eat carbs now!”

So it’s a bit of an apples & oranges comparison. The DIY is less ambitious in what it’s trying to do but that actually makes it more interesting to me. One of my big concerns about the high end ones has been “Who’s going to be able to afford one of these? What insurance plans are going to cover it and who are they going to authorize?” Those questions never come up in the “great new technology will save all T1’s forever!” coverage out there. I think there’s a HUGE gap there and there’s going to be a need for something like this DIY version. Lots of people need a car but not everyone can afford a Ferrari.

Re the danger and CGM accuracy concerns: the fudge factor in finger stick BG is pretty big too, let’s not forget. So the only real way to know how well it works is the empirical approach: strap it on and try it. So far the results of their testing look pretty encouraging.


Exactly. I can only imagine how expensive a stable glucagon product will be. Not to mention little gotchas such as interstitial fluid glucose sensors just aren’t all that accurate. Perhaps a “minority report” version of them (odd man out data gets discarded) would help, but think of the cost, complexity, and additional hardware embedded in our bods.

I met Dana at ADA in June in Boston after she spoke-- nothing they are doing is illegal. They are speaking with the FDA about it as well. You can do this in your own family. So we did it in ours-- my husband has been working on it non-stop since June and it is now reliably starting to work, in Nov-Dec, for both my son and I. It is a lot of work and you need to know what you are doing, but he was able to figure it out.

It does not bolus insulin but it changes basal rates, based on a cgms. So it still has all the limitations of pharmaceutical insulin, and you still need to manually bolus for meals and deal with exercise yourself. But for night time, it is unbelievable. I am now getting uninterrupted sleep much more often than before!

I wrote a blog about my experiences here:

Any questions let me know! It is very cool. The best sources of info are: and following #openaps on twitter.