DIY Pancreas?

People suffering from diabetes have turned to sophisticated do-it-yourself technologies. Here’s the abstract to an excellent article on these developments by Crabtree, McLay and Wilmot:

Diabetes technology has been advancing rapidly over recent years. While some of this is driven by medical technology companies, a lot of the driving force for these developments comes from people living with diabetes (#WeAreNotWaiting) who have developed their own ‘do-it-yourself’ artificial pancreas systems (DIY APS) using continuous glucose monitoring, insulin pumps and smartphone technology to run algorithms shared freely with the intent of improving quality of life and glycaemic control. Existing evidence, although observational, seems promising but more robust data are required to establish the safety and outcomes. This is unregulated technology and the off-label use of interstitial glucose monitors and insulin pumps can be disconcerting for people living with diabetes, health care professionals, organisations, and diabetes technology companies alike.

Here we discuss the principles of DIY APS, the outcomes observed so far and the feedback from users, and debate the ethical issues which arise before looking to the future and newer technologies on the horizon.

Hat tip: Dennis Sheehan.

Comments

This new technology, such as it is, creates focus, in this case focus on one's level of insulin, just as a Fitbit creates focus on vital signs (heart rate, etc.). The one thing about so-called smart devices (the smart phone, Fitbit, etc.) is that they make us obsessive compulsive. Now, being obsessive compulsive might create more focus on our health, but being obsessive compulsive will also make one miserable, which can cause an early death.

You might wish to delete that astoundingly ignorant and even hurtful comment.

These devices are overwhelmingly used by people with type 1 diabetes, for whom a focus on insulin levels creates the possibility of staying alive.

Respond

Add Comment

Yeah, what the hell, rayward. Tight blood glucose control is the single best method for improving long-term outcomes in type 1 diabetics. I hope you are never cursed with this disease, which only ever gets better when the patient becomes obsessed with blood glucose management.

I read the comment as criticizing Fitbit induced obsessive exercise, not keeping your diabetes under control. Seems off topic but not mean spirited

Not at all. While their may be a few percent who this can really help make no mistake their will be a significant percent that obsesses needlessly over this.

I am very reluctant to be rude like this but seriously, do you have any experience whatsoever with T1D, or this technology? Do you have even the slightest idea what you are talking about?

Respond

Add Comment

>"obsesses needlessly"
did you comprehend the article? connecting the pump to the cgs eliminates the need for daily manual monitoring.

How is this a risk for obsession? it seems to me like quite the opposite: handing over the monitoring to a system, & trusting that system.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Please consider deleting this comment.

Being obsessive compulsive is the only way a person with Type 1 Diabetes can hope to avoid serious complications after ~30 years of the condition. People with Type 1 track their carbohydrate intake and obsess over the food they eat so they don't go blind and/or lose a limb at the age of 55. To suggest that these people are making the incorrect decision indicates that you have virtually no idea what you are talking about, and hurts the cause of people who actually suffer from the condition.

No need to delete. If Alex deleted every stupid comment he'd have no time to put up posts for people to make stupid comments on.

Whatever Rayward intended to communicate, people have pointed out that obsessing over blood sugar levels is the only way for people with type one diabetes to stay healthy. So no one is likely to read the comments here and respond with, "Cool! No more blood sugar level testing for me!"

Respond

Add Comment

Respond

Add Comment

"This new technology..."

Not new technology.

It's just the GOP right wing, conservative view that profits must be forced to increase to create jobs by paying fewer workers less (profits is the money NOT paid to workers, but to monopolists), has created such scarcity of old technology that workers are bypassing the corporations.

The technology was developed by Dean Kamen when he was a kid, decades ago.

A China manufacturer that produced the hardware under contract is selling the hardware to end users. Obviously the high profit corporations have a huge incentive to prevent these sales by any means possible. Buying government police action is one way.

The problem would not exist if the hardware and software were sold at costs plus normal ROIC, eg 2% over prime debt given these are commodity devices. Like medical products back before Reagan.

Wtf?

mulp is basically crazy. Just ignore him.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

I have Type I, and have been following this (r)evolution for awhile now. A child’s 18yo classmate also has it, and he and his father have, using their iPhone, developed a system that constantly monitors the child’s blood sugar, allows adjustment of his insulin pump, and even allows both parents to monitor blood sugar level from wherever they are (important if the child gets low blood sugar and becomes to confused to know what is going on).

In addition, even these sophisticated innovations can be closely approximated with a standard insulin pump, and products to monitor blood sugar as often as you want available at the local pharmacy.

Not a cure by a long shot alas, but major breakthroughs.

So basically, the app is constantly measuring your blood glucose, and based on that reading your insulin pump injects an appropriate amount of insulin? Is it all completely automated? Very fascinating topic and its always best to hear directly from people with the disease.

Exactly that. A technological miracle.

Respond

Add Comment

There are multiple apps. These articles are older, but describe the process:

https://www.bbc.com/news/technology-36711994

https://www.al.com/news/huntsville/2017/05/daniel_lewis_built_her_own_art.html

Even MR had something about it earlier:

https://marginalrevolution.com/marginalrevolution/2016/05/i-medical-device.html#comments

Yes, that was my reaction upon reading this: I'd read about these DIY pumps before, probably due to a link from MR. Which as you've shown is indeed the case.

The only slight surprise is that old post came from Tyler rather than Alex, who in the past has been eager to pump out posts about alleged medical breakthroughs.

But this is not one of Alex's gee whiz posts about vaporware, it's legitimately intriguing and almost certainly helpful technological development.

Respond

Add Comment

Respond

Add Comment

Kind of - there are also other inputs that the system cannot fathom on its own, in particular carbs intake and also excercise, high temperature etc. The problem is that the system has way too much delay to work only on feedback from the blood glucose level like your pancreas does.

What do you mean when you say that is a problem? Do you mean that an as- yet-unattained ideal state exists or that there is a practical reason not to use, and celebrate, this technology?

I mean that it would be easier if it could rely only on feedback.

Wish granted. Whilst it works much better with user input of carbs intake, it copes surprisingly well with no input at all.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

This tends to show how impoverished our global medical R&D program is outside of the rent seeking world of pharmaceuticals.

Actually, the main problem is that there is only one continuous glucose monitor that is currently FDA approved for insulin dosing. Medical device technology companies would happily release many different closed-loop systems if the FDA would approve more than one CGM device for that purpose.

It's hard to think of a clearer condemnation of excess regulation in pharma than this.

The underlying glucose monitoring and insulin delivery technology has been around for decades. The benefits of tighter management are very well understood. Using a smartphone and an automated algorithm to do this better is only a small incremental step forward...yet it has been delayed for years.

Respond

Add Comment

Yeah, it's not just pharmaceuticals where you get the rent seeking. When I was diagnosed with sleep apnea about ten years ago, my first CPAP machine cost me $2,500. When it broke about four years ago, I bought one from China in a, shall we say, gray market transaction for about $500, and it it works much better than my old one.

Yeah, you can get a CPAP machine in Australia for $500 US dollars or less. I don't know how good they are though. But there's still a lot of people being overcharged. For example, being charged a fortune for a relatively small lithium battery for use in blackouts. My dad just uses a car battery.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Wait, there is actually something better than the American pharma industry when it comes to finding solutions to problems that people confront?

Only a cynic would suggest that the pharma industry will be more than happy to have the FDA be their front for profit protection.

I get indignant when Alex criticizes the FDA, but I also believe they're a corrupt organization that's been captured by the industry they're supposed to be regulating. You figure this out, because I can't!

Surely you meant current dissonance.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Hope this DIY technology is not misused to transfer liability to others, e.g. "don't worry teacher, Timmy is now wearing my trinket and he'll get the right dose at any time at school", or "yes teacher, I don't need to take care of Timmy during the school trip, I can see his glucose level on my phone. it's better if I'm not there......socialization with other kids it's important, its not that I have better things to do".

Believers can believer whatever they want, the problem starts when they proselytize.

Disclosure: my wife's a teacher and the integration of children with health problems into normal classrooms is a challenge. Is the teacher a nurse or a teacher?

Teachers cannot manage this (T1D), but this technology removes all but the most remote risk of any teacher management being required. So seems a clear win on that front.

Respond

Add Comment

Respond

Add Comment

News Blackout. Supportere of Iowa putschists are deleting all denunciations of the Buttilieg-led coup in Iowa. The Iowan junta refuses to release the real results. It is time to stand and resist. We demand free and fair elections in America.

Respond

Add Comment

"We demand free and fair elections in America."

So do we.

My point is, a recounting is in order. Iowa has become this generation's WACO.

Respond

Add Comment

Respond

Add Comment

There's no mention of them using a PID loop for control. If not, then it's not going to work very well. Granted, it's possible that the author of the article didn't understand the control system very well and just left it out.

But you should be using a PID loop (or at least a PI loop) for this kind of control. It's far more accurate and won't overshoot nearly as badly as a simple I/O control would.

PID = Proportional Integral Derivative
https://en.wikipedia.org/wiki/PID_controller

In layman's terms, a normal control loop tends to oscillate around the setpoint. A PID loop quickly damps out the oscillation so that the process value remains close to the desired value.

It has worked out for years now, I guess they must have thought of that :)

Respond

Add Comment

Umm, PID control isn't exactly new; as far as I know, it's been around at least since 1920 or so, and thus pre-dates digital technology. I can't imagine anyone working on a closed-loop system without some knowledge of it.

Nonetheless, the proportional term remains by far the most important in practically all such systems. And the integral term isn't there to prevent overshoot (which results from too much gain in the feedback loop combined with the inevitable time delays) but to prevent errors from persisting. Because, integrating even a small error over time will cause the integral term to grow ever-larger, thus forcing a correction if the system is working at all.

Aside from that, in a digital controls system you want to be sure everything is initialized so that bizarre stuff doesn't happen at startup. And, you'll probably want some sort of independent supervisor system to prevent disaster if something goes wrong, such as a failed sensor. I.e., you wouldn't want your system to put you into insulin shock if its glucose sensor under-reports your blood glucose. Just as you wouldn't want an aircraft control system to endanger the aircraft if a single sensor (such as an angle-of-attack sensor) were to fail.

At least, I assume engineers at Boeing Corp. would agree?

"Umm, PID control isn't exactly new"

I wasn't saying PID's were new. I was saying they are basic control technology and yet the article showed some code examples that were just simple If-Then statements.

More than likely the people doing the programming are using some kind of defined PID loop but it was lost in the abstraction of creating an article.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Diabetics are waiting for a biological cure.

Respond

Add Comment

Why read an article when you can read what someone actually involved in the work writes?

The link is https://www.hanselman.com/blog/ThisChangesEverythingForTheDIYDiabetesCommunityTidePoolPartnersWithMedtronicAndDexcom.aspx

Respond

Add Comment

Bloomberg Businessweek ran a feature on this topic in 2018...

https://www.bloomberg.com/news/features/2018-08-08/the-250-biohack-that-s-revolutionizing-life-with-diabetes

Respond

Add Comment

Do the devices allow users to use the cheap insulin?

Technically, yes, although the benefit of this is greatly reduced. Very rapid-acting (expensive) insulin can take as much as 10 minutes to react in the body. Older, cheaper insulin reacts on a longer time horizon, and thus the lag between the reaction time and the CGM's blood glucose reading is far enough apart that I'm not totally sure there would be any advantage to this.

I'm on MDIs, though. Perhaps someone with more direct experience can weigh in.

Thanks for the answer.

Respond

Add Comment

Respond

Add Comment

They do not really work on old insulin due to the lag: they work, but much less than with eg Fiasp. But this is an American problem only.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment