As opposed to not doing anything which is the current state of Diabetes for the most part, which too, kills you and if not outright kills you then slowly does via financial ruin.

How many episodes of hypoglycemia do you think one needs to have, in an utterly broken healthcare system in which being ordered paramedics and given sugar pills costs you $XXXX per occurance.

And if you're gonna talk to me about responsbility and healthy living please join the real world some time.

That's why people choose to take actions into their own hands. Because being a victim of a system that just doesn't care enough leads you to either take the handles or down under.

What does this even mean? Hacking an open-loop insulin pump and making it closed-loop doesn't eliminate the need for insulin, did you misread me as saying insulin prices are A-OK and diabetics should just eat healthy?

All power to these knowledgeable, individuals, for doing something dangerous but potentially great. This is HN, full of smart people who understand the risk of failure and will monitor the crap out of their system and adjust and tinker until it's ok. The problem is people making tutorials and publicizing their work, which leads to people not as smart following their tutorial and thinking it will be alright.

There are tons of papers on closed-loop systems, and probably clinical trials are happening as I write this comment. Medicine isn't one of the industries to "disrupt" though, as much as people on here want to believe.

Smart insulin pump sounds awesome. Bring them on. If it really is a smart insulin pump. Trouble was the article author was talking about 'reading his emotions' etc. There's a red flag for bullshit. AFAIK no publicly disclosed technology can, in general, read emotions. Also, why whatever signal they are using as 'emotions' is a perfect predictor of insulin levels, I don't know.

I studied control engineering, and I bet he forgot some things. A few things come to mind. How does he recognize sensor failure (pretty important in closed-loop systems, like MCAS for example)? Did he use floating point math which may cause numerical issues? Does he use integrators and limit their output? If he does limit the output does he counteract windup? If he uses neural networks, did he gather enough data that we can be sure that this things works?

There's so many pitfalls here where teams of engineers with the whole part spec and from different disciplines have failed before. It sounds very dangerous.

Fuzzing seems like a really useful test for this.

Either way, I agree I'm not really subscribed to the idea that DIY life critical systems are a good idea beyond POC.

Maybe it isn't life critical but only quality of life critical?

If it pumps too much you're dead - which is why I'm considering it life critical.

Not enough is also bad but hopefully they'll notice and can act accordingly - still a threat too though.

It's not that we're all saying this is an unsafe machine, we're saying that a random person DIY'ing and using such a solution is very risky, there's a reason why commercial medical equipment requires rigorous testing/certifications, etc.

Insulin pumps are both.

Usage of cheaper slow acting insulin becomes viable if it is controlled by an algorithm that knows your insulin level.

You can already use slow-acting insulin, it just requires planning your day out really well, which is hard and why slow-acting stuff sucks. The amount of data required to predict someone's glucose consumption/insulin requirements is massive, and needs to come from that individual. Then you'd need to build in overrides because people change their habits/behave spontaneously which puts them in danger. It isn't a simple problem.

That doesn't address his point, which is that a programmable pump isn't something you could sell because of the liability involved.

It is something you can sell, and is currently in development by major pharmaceutical companies.

Not end-user programmable, though.

I think the parent did not criticize the DIY community, only stating that different rules apply to biotech companies. I know many people who use the loop system and their health benefits immensely from it.

I will just say this. Someone in the openloop / DIY pump community is eventually going to die or seriously injure themselves (or a family member) doing what they do. It's not an if , but a when.

As long as this is someone who's actually working on the tech and knows what they're doing, this is only an unfortunate and very sad cost of progress. It's the risk they take, because they know they're doing good work.

This does become a serious problem when it gets into the hands of people who don't know what they're doing, if they're convinced to trust the tech because "obviously the smart people working on it made sure it's safe".

Still, parent comment stands. We might agree it would be better to do something, but regulation is regulation. To change that, you need to vote the right pers... Ouch, I guess that never works either.

While I understand the issue in general it seems it should be relatively simple to validate that a closed-loop insulin pump is at least no worse. Isn't the objective of the closed-loop to just do finer control? So can't you just have the same hard limits as the normal pump and then allow the potentially buggy software to just fine-tune within a generally safe range?

Lot's of things can kill us. People have been changing their car brakes for a long time. If your car breaks fails, it can kill you, your passengers and other's outside of your car. People have been fixing their electrical appliances, it can electrocute you. Gas lines and furnace can explode. There's a lot that we do that can kill us if done incorrectly. This doesn't mean that we shouldn't do it. At the end of the day it's humans that built it.

I see your point and agree... there should be no difference. But I think a lot of times it's not WYSIWYG when it comes to software, because there is so much hidden infrastructure on top of something simple.

I don't understand the justification. Because people do dumb things, we should do more dumb things?

>This doesn't mean that we shouldn't do it.

Why should she put her son at risk for a science experiment?

Because the risks of not doing so are even worse.

If it seems horrible, that's because the reality that diabetics have to live with is horrible.

Are you saying that changing your own car brakes and fixing your own electrical appliances are “dumb things” to do?

Crossing the road turns out to be pretty dumb too, let's just stay at home in future.

To be fair that same process allowed hackable (malicious or not) devices out into the world

https://www.webmd.com/diabetes/news/20190628/fda-recalls-ins...

Believe me I'm not saying the FDA is some perfect organization that can do no wrong but they do exist for a reason. There definitely needs to be more scrutiny as tech (read:computers/AI) start making their way into therapeutic devices.

It's a trade-off. You can choose to wait for something to come to the market and suffer the potential consequences, or you can go the DIY route and reap the benefits today. There are risks either way. My wife and I aren't willing to experiment with our son's life. If it were for myself, I'd absolutely give OpenAPS a shot. My son has a Dexcom & a Tandem x2 pump, and in Q4 they'll deliver a software update that will provide a basic closed-loop system. 'Basic' in the sense that it will only be able to reduce the likelihood of a high blood sugar. We're still waiting for the dual-hormone pumps that will be able to control blood sugar in both directions.

Prior to this pump + Dexcom, he had a Medtronic 530g + glucose monitor. It was not internet connected out of the box, it required a $100 gadget that would transmit his blood sugar to his cell phone and then to medtronic's awful website. Fortunately folks developed Nightscout to scrape Medtronic's website and provide a much better experience. With Dexcom being a much more modern option (the CGM transmitter has built-in bluetooth) - we don't really need Nightscout, but it's convenient for me to watch his sugar on my 2nd monitor when he's at school (he's only 7.)

The problem is not in my opinion regulatory processes or clinical testing. It's been simply a lack of urgency/competition. Beta Bionics has started to make Medtronic and other competitors step up their game in recent years. https://www.betabionics.com/

The situation reminds me of the regulatory environment in aviation. Aviation in general is super regulated for the usual “people could die” set of reasons. If you buy a normal certificated airplane like one from Cessna or Piper, you are subject to heavy regulations around what you can do to it in terms of maintenance. You can’t do much as run a USB charger to the cockpit panel without the FAA coming down on you like a ton of bricks. You have to go to people with the right government $tamp to get anything fixed. It’s a huge burden but it’s all in the name of safety.

The FAA does however have a parallel track, called Experimental-Amateur Built (E-AB) where if you build the airplane yourself for educational or recreational purposes, it’s subject to a mostly separate set of rules. They are generally more relaxed in terms of what maintenance and modifications you can do but stricter in terms of what you can do while flying (can’t take passengers for hire, for example). Most pilots prefer the certified/safety-first/highly regulated track, but the government recognizes it’s a trade-off and offers the other option for those who prefer a more DIY-friendly world.

Not saying this would necessarily work for medical devices, but there is precedent for acknowledging the trade-offs when it comes to DIYing safety-critical stuff.

It's understandable that the regulatory process errs way on the side of restrictiveness: if someone dies because it's not restrictive enough, that's easy to identify, and it can become a lawsuit and a political campaign, Senators can get involved, agency heads can resign, civil servants can get fired. Meanwhile, the people who die because it's too restrictive are much more difficult to identify, and their deaths can't be blamed on a single decision. (Maybe the unapproved treatment wouldn't have worked anyway, and maybe nobody even knows it could have existed.)

So, given that whatever decision they make will kill masses of people, regulators have very strong incentives to kill them by being too restrictive, rather than kill a different mass of people by not being restrictive enough.

The solution to this problem is clear: the person with the strongest incentive to keep you alive is you. Any regulator or drug-company executive will reliably put other considerations above your survival under some circumstances. So, except in cases of people who really cannot be trusted with life-and-death decisions — children and some severely mentally impaired people — you should have the final say on what medical treatments you try or don't try.

In short, the whole regulatory structure should be scrapped. We still need clinical trials, but not before new treatments become available; after.

This sounds like a radical proposal, but in fact it's how the process worked in the US during the time that we owe most of our revolutionary medical treatments to, the 1920s through the 1960s. That's when most (all?) of our currently marketed antibiotic families were discovered, when one dude spent a couple of years prototyping the pacemaker in his barn and then convinced his friend at a hospital to start trying it in patients, when artificial joints were developed. We're still coasting on the legacy from that era.

> I work in Biotech and every time I read one of these articles/posts about how somebody hacked their insulin pump people start asking why there aren't already affordable closed-loop pumps on the market if people are hacking their own with raspberry pis and the like, and they clearly don't understand the regulatory process or the level of clinical testing that's needed to validate something that, if it malfunctions at night, can kill you.

plenty of closed loop feedback systems we are all hooked up to which can malfunction and kill you.. casually thinking of automotive industry I can come up with:

- abs brakes - engine control units - transmission control units - power steering - traffic lights

and that's just one sector, let alone aviation, mass transit, power plants, etc etc etc.

granted the malfunction:fatality ratio is probably different, but not so altogether crazy to think that 'closed loop is impossible' .. though perhaps you are strictly speaking w/r/t regulatory climate..

that said, we have had pacemakers for many years..

The alternative to a pacemaker is death from an irregular heartbeat. The alternative to closed loop insulin is a lower quality of life.

The FDA is setup to optimize survival not quality of life. The standards will be much higher for these pumps.

Yes makes me shudder thinking of what could go wrong .

I spent a year hooked up for 8 hours a night to a CPD dialysis machine - I read the massive manual and there is a long section on what can go wrong including "Death" - no way would I start hacking on that.

Though I did think about using a camera and ML to monitor the waste output for signs of contamination.

And yet MANY people willingly make that tradeoff. https://www.usatoday.com/story/tech/2019/06/05/diabetics-for... Also with CPAP devices, which are less likely to kill you but it's still possible. https://piunikaweb.com/2019/02/21/rise-and-fall-of-sleepyhea...

People also tend to greatly underestimate the risk of negative outcomes.

this implies that if a medical device (or pharma) passes through the regulatory process it will never malfunction which is absolutely not true.

It doesn't imply it will never malfunction. It implies it went through a scientifically rigorous process to minimize the chances it will malfunction.

And often that there's warranties in place - that is, if it does fail under normal operation, you can sue them and get millions. Which won't buy your loved one back but it's a strong deterrent for the company to not try and palm off dysfunctional shit.

that doesnt help if youre dead or severly injured...besides how many of these cases actually go to court and win?

It doesn't fix those things, but it does help deal with them, at least where the long-term financial impacts are concerned.

You're not going to squeeze a lifetime worth of lost salary out of a random Redditor who told you how to tweak your insulin pump.

It really just implies that the odds are sufficiently improved. Which may or may not be true, but which has a much better shot at it.

often times products are expedited through this ‘scientific process’ to get on the market faster. if the regulatory process is truly about safety why does this exception even exist?

Whether it's actually a good idea isn't something I know enough of the real numbers to comment on, but it seems plenty possible that an expensive step that substantially reduces risk might be appropriate in some circumstances and not in others.

> if the regulatory process is truly about safety why does this exception even exist?

Because the risk is worthwhile at times.

The exceptions typically apply to rare diseases that impact a population too small to be profitable enough to justify the $1B or more it may cost to shepherd a device or medication through the normal process.

These are populations who would be potentially unserved otherwise. (I'm of the personal opinion that this is where the government should step in instead of private industry, but I'm pretty left of center.)

https://www.fda.gov/industry/developing-products-rare-diseas...

IIRC, an exception of this sort was used to expedite an Ebola vaccine due to the recent outbreak... because, well, Ebola.