I've been using this system for ~2 years and it has completely changed my life (not an overstatement).
About 6 months ago I realized that as more and more people found out about this (and Atlantic publishes articles about it...) these pumps were going to be harder to find. I immediately scoured craigslists across the county and spent an inordinate amount of money to buy three of these pumps, because once they break, it's game over.
I can't imagine having to go back to the old way, it actually scares me.
This was on my mind the entire time I read the article. People are having their lives opened up by responsive pumps; some of them are even children who will grow up without ever having to manually manage each blood sugar swing. But the entire setup depends on scavenging outdated pumps and importing unapproved devices. It's deeply scary and depressing to think of where they'll be left if those pumps fail before an approved solution hits the market - or becomes similarly affordable.
Even if closed-loop pumps are approved and affordable at that point, the loss of control will represent a real step backwards for a lot of people. I don't have easy answers, but realizing that this is how things work is chilling.
What kind of effort do you think it would take to make an open-source functional copy of one that you would trust? Presumably if doing that the reverse engineering may have to be black-boxed and the designs presented as research, to create some legal cover.
Honestly, I think this should be quite doable. These guys [1] built a pump based on a stepper motor and an arduino, in order to replace expensive peristaltic laboratory pumps.
It's an ongoing reminder that the Silicon Valley point of view isn't the only—or best—one.
I grew up dirt poor in the Deep South. My wife is diabetic. People need the perspective. If that's flamebait, maybe I'm not the one who's in the wrong.
Also it seems like you're ignoring your own "don't post shallow dismissals" rule.
To do that though we'd need to solve the problems of manufacturing bandwidth (for-profit drug manufacturing as it exists now wouldn't work in a patent-free world) and drug discovery (a large amount of which occurs in for-profit environments that seek to file patents). How do you address that in a pure-capitalism system? Drug discovery is incredibly expensive.
My personal preference would be to set a hard time limit on drug patents (insulin is so old that it should have gone generic decades ago, but the drug companies cheat) and ban the government & universities from selling patents/derived patents to for-profit companies (insulin, for example, was one of these cases). A hard time limit makes it possible to manufacture things like generic insulin but companies can still keep developing and monetizing new treatments. I don't think that would probably fix things in the long term though, they'd find ways around it. Banning patents from leaving government/university hands would mean more new treatments would be developed using public funding and then any company could manufacture them and compete on the merits.
There are so many opportunities for drug companies and medical care organizations in general to cheat patients - you don't have the freedom to shop around for alternative prices and quotes if you need an expensive therapy Right Now so that your liver doesn't shut down. How can you protect people in that position just using the Free Market? There have to be regulations of some kind, so which ones?
One popular suggestion lately is to just nationalize health care, because the government can tell pharmaceutical companies what it's willing to pay and if they want to charge 5000% more they can pound sand. It's my understanding that some countries have also just started ignoring patents and manufacturing their own drugs, which seems like another option. Could either of those solve it here, though? Who knows. Likely not as long as lawmakers love drug companies and patents.
Manufacturing is commoditized. The bandwidth exists in the generics factories of India. They're able to retool fast and get out high quality stuff pretty fast.
Discovery is a problem, but GlaxoSmithKline is only $200 billion dollars. The government can easily afford drug discovery grants.
Therapy that involves human beings is tough and I don't think I have an answer. But drugs are different. If I can buy cheap generics from India, I should be able to. If I want, I should be allowed to resell. If the government wants to say "FDA-certified" on some, so be it. For the rest, let the people decide.
By acting like a parody of an evil capitalist combined with a youtuber's penchant for stupid stunts, he's made a lot of people now believe (not without foundation) that all drug pricing is an arbitrary ripoff. He is now in jail for a mostly unrelated securities fraud.
I said insulin and meant insulin. Insulin prices have skyrocketed over the last decades beyond any point of comparison without any significant change in the quality or manufacturing process.
As you say though, this is not true of all drug pricing. However, given we're talking about insulin pumps for diabetics, it's reasonable to assume they'd be priced based on the same philosophy as insulin, no?
Shkreli is a gaudy distraction, the same way Jordan Belfort was a gaudy distraction from much quieter fraud and mislabeled risk.
You're absolutely right to focus on insulin. In addition to the obvious relevance to diabetes pumps, it's the perfect example of how pricing isn't based on high costs of drug discovery or manufacturing. The basic market pattern people expect still holds: newer formulations cost more than older ones. It's just that the older formulations never get cheaper; contrary to all economic logic, their (constant-dollar) prices have risen over time. That doesn't bode well for any other step in the diabetes management process.
Shkreli is just the only one dumb enough to say out loud what other pharma executives keep to themselves. His drug pricing strategy is not unique, unfortunately.
Capitalists created all of the pieces of technology in use here. Bashing them seems really out of line.
Getting devices approved by the FDA is extremely expensive and time consuming. The DIY guys are doing great work, but they have not done this. An actual company such as Medtronic could not do what they have done as they would get shut down by the FDA.
The US is the only industrialized country that does not have the English Rule, which says that if you sue someone in court and lose, you pay their court costs. As a result, trial lawyers make a lot of hay suing medical companies. There's no strong disincentive to see what happens if you go to court. This is another reason medical companies are very conservative when releasing equipment for use.
So, we have capitalists producing technology despite all of these hurdles, and yet somehow you blame them when they take a while to get things into the market?
Contrary to your belief, the entire asset of technology known to mankind was not all invented due to capitalism. We had a lot of political systems throughout our past.
Patents, and copyright, are only in existence because of government's law. Without such, you could still live in a perfectly capable society (with capitalism as its political system, but not necessarily).
The pumps and monitors and smart phones being used by the DIYers in this case were all produced by capitalists. Without these things, the DIYers would not be able to do what they are doing.
The capitalists got these devices to market despite the regulatory processes of the FDA, which are expensive and time consuming.
What this story highlights is that the regulatory process has created a situation where only corporations with big pockets can generally play in this space legally. What the DIYers are doing is not legal. It is definitely moral, but not legal.
I'm working with a group trying to put some dent in this problem called Open Insulin (openinsulin.org). It's very legally uncertain/risky however, since the current regulatory regime is impenetrable. The problem here is not technical, its legal -- we have the technology to make this type of stuff cheaply and safely, but the status quo will not let us sell these things products without going through a multi-year, multi-million dollar regulatory process fraught with arbitrary constraints and risk of multi-million dollar patent lawsuits. Until then, the best we can hope for is feeding off scraps via loopholes, as you're forced to do.
I vote you do it underground and illegally until the law catches up. "If a law is unjust, a man is not only right to disobey it, he is obligated to do so." - Thomas Jefferson
That said if I had Type 1 diabetes and the skills I absolutely would do it.
Given the amazing things open source software and hardware folks have done and that you can manufacture small run electronics easier than ever it seems like a healthy underground is likely to emerge.
I'm in favour of the FDA and organisations generally (otherwise we are back to snake oil again) but it has massive issues with technology, they simply can't move fast enough to keep up.
I think a pro-tem solution is a law allowing a patient to waive all rights to sue in return for been able to use a device.
That way technology would get into the hands of the people who need it, the FDA would have a better understanding of what is going on and oversight and small manufacturers can enter the market.
Drug trials to an extent already have such legal frameworks present otherwise we wouldn't be able to perform the trials.
> I think a pro-tem solution is a law allowing a patient to waive all rights to sue in return for been able to use a device.
My proposed solution is to make the FDA optional. If somebody elects to use a non-FDA approved device or substance, that's their decision to make -- not a group of faceless bureaucrats in Washington D.C. New regulatory institutions should be permitted to form, and so long as they're transparent about their process, and sellers are honest about who they've been audited b, false advertising and other species of fraud can be dealt with in the courts. People ought to be free to choose for themselves which institutions to trust and which to be skeptical of.
As soon as an institution becomes involuntary, it becomes immune to rational inquiry and revision. "Because we said so" is the ultimate axiom of any idea enforced by the FDA - if any individual disagrees with their thinking, well too bad you have to do as they say, regardless. Independent thought is the foundation of all truth, so its not really a mystery why domains where its illegal to disagree with the status quo, like medical regulation, are so fucked up.
> I think a pro-tem solution is a law allowing a patient to waive all rights to sue in return for been able to use a device.
I think we arrived at the same conclusion from opposite ends.
The big advantage to the FDA for the drug companies is that if they get FDA approval they top cover from that when they get sued.
Also the cost and scope of getting through the FDA makes it harder for smaller companies to get to market without either selling out to the larger companies or merging which also favours the larger companies.
It's an interesting dynamic, we definitely need some oversight of what is in the market but no so much that will stifle the market entirely.
It is already legal to produce your own insulin at home and take it yourself. AFAIK its even legal to give it away for free. The problem is
1. It's expensive. The final production setup is estimated at $10k
2. The R&D non-trivial. We're still building the platform.
3. Production still must be highly self-regulated to ensure that the insulin is safe.
People need a way to fund this sort of work, but the second we try to sell anything we need to go through a multi-year, multi-million dollar FDA approval process or we'll be shut down. We're currently exploring other funding models, certainly open to ideas.
That seems like a great kickstarter you should make then.
To avoid the need to go through FDA approval, don't label it for humans. Label it for animals. Label it for lab testing.
Buy sensors from reputable names, plug them to a chip that support uPython, and make it hackable.
Even better, you may be able to recruit an engineer from one of the original providers to secure the pumping system, just for the sheer human aspect of it. Geeks are often moral driven persons.
Soon you'll have an affordable, community driven device.
There will be problems, and some people will probably pay it with their life thought. But how many lifes can be made better ?
I imagine that for a lot of people "oh well a few people died, but it's for the greater good!" is in fact not enough justification to help them sleep at night.
I get that, that's why you don't label it for humans, and you make it hackable. Just like there are people ready to die on Mars just for the sake of making the next step for humanity, there is out there a community of people ready to take the risk with their body to solve a problem that plagued their community for so long. Who knows, they may actually want it, and not be able to do it because of the lack of opportunity.
Open the door, let people volunteer. If you get a response, the ones coming will know what they are in for, and as long as you are respecting them, their action, and the community, you can be proud of what you do.
Why do you think Pasteur tested the vaccine on himself ?
The desire to help, being part of something important, is a deep human drive.
That's my point. They may be people like him for a new type of insuline pump. We can't, as citizen, hope to compete against pharmaceutical companies. That doesn't mean we can't do anything.
> They may be people like him for a new type of insuline pump.
I'm sure there are -- and there is nothing stopping those people from doing the same today. The FDA cannot tell you what you can and can't do to your own body.
Since this is such a mission critical piece of hardware that would severely impair the user if a bug or malfunction developed, how can you completely trust the device will continue to work 99.9% of the time? I really want to recommend this to loved ones but hesitate with the liability of over/underdosing insulin and them not being as tech saavy as the HN crowd...
To answer the question you're asking, I trust that the _device_ is basically never going to fail because it's FDA approved and manufactured by Medtronic, who has a loooot on the line to make their devices fail-safe.
To answer the question I _think_ you're getting at... I trust the open source software that runs the FDA approved device because it's designed in such a way as to be as careful as possible (temp basals vs boluses). It's also open source and a lot of people contribute to it. I also only update it once a year or so when my Apple Developer certificate needs renewal.
All of that ^ helps, but honestly, it's not a guarantee. I still wear an off the shelf, FDA approved CGM with their native 1st party app installed so I still get their alerts on low/high blood sugar.
To me, it's a calculated risk and one that (so far) has been extremely worth it. My A1C went from ~7.5 to ~5.6 after using it.
I feel the same, however, people not having access to such an automated insulin delivery system may also suffer/die because of the consequences of a worse regulated sugar.
Of course it is a calculated risk, however as a T1D myself, you can't ignore the fact that the current FDA approved pumps are also very dangerous. Pumps that don't respond to dropping blood sugars (most of them on the market today) and keep delivering insulin when you don't need it, compound low blood sugars, which has personally got me into serious trouble many times. Picking between a FDA approved "dumb" pump running on 10 year old software versus an open source option that explicitly has code to stop insulin delivery when blood sugars fall, seems to be an obvious choice. I've had diabetes for 31 years and this is by far, hands down, the greatest advancement I've experienced.
A valid concern, but also humans are guaranteed to make mistakes and I'm sure a high percentage of them, so anything that lowers that error rate is an improvement even if it's not perfect. My thoughts anyway
In the business this trust comes from the fact that they are regulated, validated devices (lots of testing and process; any change involves a lot of work). I don't think the DIY movement will address regulation or validation (and I don't know that regulation and validation is guaranteed to make a safer product than DIY can).
Both Humans and Software can malfunction. The hardware is identical. Also if the algorithm does small injections every few minutes then the risk of overdosing/underdosing is extremely low compared to a single big injection at midnight.
In Germany we have four pumps available for looping with AndroidAPS:
Roche Accu-Chek Combo and Insight
Sooil Dana R and Dana RS
From these Sooil is the only one who allows looping. Been using one for a month now and there is no going back.
I was using a used Combo last year but it broke suddenly. This time I have four years of warranty and hopefully we have more models available at that point.
I was under the impression that there were already devices on the market that would monitor your blood sugar and dispense insulin as needed. Is that not the case? Or does this system provide some additional functionality?
There is one pump that offers looping. However, it mandates a blood sugar floor that is 1.5 times the non-diabetic. The glucose monitoring does not allow the user to control the alarms or snooze them. When these devices are connected to smart phones the user does not have ownership of their own data and the UI does not show all the data sent with the packets.
About 6 months ago I realized that as more and more people found out about this (and Atlantic publishes articles about it...) these pumps were going to be harder to find. I immediately scoured craigslists across the county and spent an inordinate amount of money to buy three of these pumps, because once they break, it's game over.
I can't imagine having to go back to the old way, it actually scares me.