On Chronic Illness, Proprietary Software and Not Being Dead

Image provided by Denise Chan on Flickr

One week when I was twelve years old, I got a thirst that I couldn’t quench. I got so thirsty that I could drink multiple litres of water in one sitting. I got so thirsty that I’d drink a litre of water in the bathroom while passing the last litre. I got so thirsty that I would daydream of drinking from rivers and streams during classes at school. I got so thirsty I started waking up multiple times a night to drink water. I was so preoccupied with drinking water that it took over my life. I lost a little over half a stone (around 4 kilos) in a week, with no real effort, and my eyesight started to blur.

I went to see a doctor, who took a fingerprick blood glucose test, which was too high for the handheld testing machine in the clinic to register. I was promptly admitted to hospital. A normal blood glucose is around 4–8 mmol/l — mine was around 42 (most measuring equipment functionally maxes out around 25).

That evening I was diagnosed with Type 1 Diabetes (hereafter T1D), an idiopathic autoimmune condition whereby my immune system has destroyed my pancreas’s ability to produce Insulin. Without Insulin, the body cannot metabolise glucose in the blood, a process used to convert carbohydrates into energy. This is a crucial homeostatic process — without any insulin the body has to metabolise fat directly for energy, a wasteful process which produces acidic Ketones as a byproduct. With Ketones leeching into the blood Diabetic Ketoacidosis then develops, with potentially lethal consequences. Without insulin, T1D is invariably a death sentence.

In addition to having no clear cause, T1D currently has no cure. A cure is always theoretically on the horizon, however that horizon is still practically receding for diabetics as time moves forward. The treatment for T1D is daily, regular insulin injections, manually replacing the insulin the body normally produces with injected insulin to attempt to maintain normal blood sugar levels. However, the solution isn’t as simple as performing a regular injection for ‘like-for-like’ treatment. Because blood sugar is mainly affected by diet and the types of food consumed, insulin doses have to be calculated based on the amount of carbohydrates in food (however this can be greatly affected by other components of food such as fat). In addition to this, blood glucose levels can be affected by physical activity, stress, adrenaline, physical and mental illness, drugs and medications, menstruation, pregnancy, dehydration, insulin condition and body weight. If doses are miscalculated the resulting effects can be dangerous — too much insulin results in hypoglycemia, which results in symptoms resembling heavy intoxication (sweating, fatigue, dizziness, weakness, hunger, confusion, anxiety), and in extreme cases can result in convulsions, loss of consciousness and coma. Not enough insulin results in hyperglycemia, which is not immediately dangerous and has less serious symptoms (acute irritability, thirst, mood changes, blurry vision). Long-term hyperglycemia (caused by poor control of blood glucose) however can cause significant long-term health complications including infections, slowed healing, amputation, vision loss, peripheral neuropathy, erectile dysfunction, stomach issues, chronic intestinal issues, blood vessel damage, kidney damage, impaired hearing and potentially Alzheimer’s disease.

It is a lifelong illness, which fundamentally alters the way in which I live my life. I am continually mediating high and low blood sugar alongside the food that I eat and the activities I perform, attempting to keep my body in check so that I don’t pass out, and I don’t destroy my peripheral nervous system. It’s a concern at most times in my life, particularly every time I eat, and is something I forget about at my peril. Controlling my T1D is something that requires an awful lot of diligence and concentration, and shows no sign of abating.

While there is no current or near-term cure for T1D, technology has advanced the ability to control blood sugar levels considerably. Insulin was only discovered in 1921, with insulin derived from animals being the only option for injection until commercially available human insulin could be manufactured by biotechnology in 1982. Insulin injections have been administered in various forms, generally getting less painful and more advanced as time goes on, ranging from painful intramuscular injection (which would cause sores and bruising) to insulin inhalers and jet injectors which don’t require needles. The methods I currently use are administering insulin via pump, and checking my blood sugar mainly using an interstitial fluid glucose monitor and sometimes a fingerprick blood glucose test. This combination of technologies makes managing my diabetes more convenient (I’d never say easy) — particularly interstitial fluid glucose monitoring which allows me to detect trends in blood glucose levels. I rely on temporary body implants coupled with computational technology to help me perform my own homoeostasis, and to not have my diabetes kill me.

With this in mind, let’s talk about another concern of mine, Free Software.

Above is a great TEDx talk by Richard Stallman, detailing the basics of free software, what it is, how it functions, and what political purpose it serves. I’d really recommend it to anyone who is interested in the politics of software.

From the Free Software Foundation Website:

Free software is software that gives you the user the freedom to share, study and modify it. We call this free software because the user is free. To use free software is to make a political and ethical choice asserting the right to learn, and share what we learn with others. Free software has become the foundation of a learning society where we share our knowledge in a way that others can build upon and enjoy.

I am an advocate of free software, and I have chosen to use free software both in my day to day computational tasks, as well as to construct the art I create. I fundamentally believe that using free software (as opposed to proprietary software) can create an ecology that supports the freedom of users, and gives me meaningful control of the technology that I use in my everyday life. Using free software for art also allows me to build a free (as in freedom) artistic practice that can (and has) benefit a community of artists outside of simply producing work myself. I try to use free software wherever possible. I use GNU+Linux and Android on my devices, and I make music and art using SuperCollider, TidalCycles, Python, Processing and others. While I’ve found it’s not practically possible to use entirely free software (particularly some low-level drivers and high-level online services), I am not controlled by the traditional Apple or Microsoft ecosystems, and I am much freer to do as I wish with my own technology.

In terms of free hardware, the problem is more complex — the FSF acknowledges that freedom in software and freedom in hardware are practically very different. It is possible to use hardware that conforms to a similar set of standards to free software, with the FSF calling for freedom at the circuit level with the distribution of schematics under copyleft licensing, however this is often not the case in the technology I use (with the exception of hardware such as Arduino). In the absence of true free hardware, an indicator of ‘freedom’ I usually defer to when talking about hardware is the Makers’ Bill of Rights, and things like the Right to Repair. Can I maintain, repair and understand the hardware? Does it use standard connectors and screws? If so, I would consider it ‘more free’, and I attempt to use the freest hardware I have available, at the very least hardware that I can maintain and repair. In this case, meaning that it doesn’t actively impede my free use of that technology by design — such as a Surface laptop ‘made of glue’, Nintendo hardware that requires nonstandard specialist screwdrivers to open, Apple phones that use proprietary connectors for charging and a $400 juicing machine that simply opens and squeezes things.

As liberating and rewarding as I find using free software for art and everyday technology, I cannot escape proprietary software and hardware when managing my T1D. I currently (as of August 2017) use an Accu-Chek® pump, Accu-Chek® blood glucose testing meter, and a FreeStyle Libre Flash Glucose Monitoring System, which are products of parent companies F. Hoffmann-La Roche AG and Abbott Laboratories respectively. All aspects of my diabetes care (with the exception of some manual injection equipment) are digital, and all aspects of my diabetes care are proprietary. I depend on proprietary hardware and software in order to not die.

Why is this a problem?

As Richard Stallman says in his TEDx talk, with proprietary software ‘The users do not control the software, the software controls the users’, and this raises some issues relating to my T1D.

The hardware I use to control my T1D is proprietary — I cannot build this technology myself, as the schematics and information used to construct the technology that keeps me alive is not available. I also cannot maintain the technology for the same reason. Diabetes UK estimates that an insulin pump has a startup cost of £2000–3000, with consumables costing £1000–2000 a year, and for those lucky enough to have state funding (myself) or those to whom this is an acceptable cost this is not an issue. However, the cost of this is prohibitive for so many. These technologies can also only be fixed by their manufacturer at great cost — self-maintaining the things that keep me alive is not an available option — it fails all applicable conditions of the Maker’s Bill of Rights. This is a situation that I find unsettling — I have no meaningful control over the technology that keeps me alive every minute of every day.

In terms of hardware, consumables are an issue too. The main cost of diabetes is not the pumps and the meters (meters are often given out free), it is infusion sets, needles, test strips and blood glucose sensors, all of which are proprietary. When I was first diagnosed and only used fingerprick blood glucose tests, I received a talk from a clinician on the cost of blood glucose testing strips for proprietary meters, and told to be mindful of the cost of my testing to the state (I’ve been given this talk a few times) — the price of test strips for proprietary meters is astronomical. In the context of austerity in the UK, this has led to test strips being rationed to ‘save money’. For the majority of diabetics in the UK, proprietary test strips are the backbone of bodily maintenance — in absence of severe symptoms which suggest immediate harm or danger, test strips are a way to determine whether you are about to start shaking, sweating and mumbling, whether you are silently destroying your body through high blood sugar, whether you need a slight insulin adjustment, or whether you are just getting on fine. If I couldn’t test my blood sugar because the state couldn’t afford very expensive proprietary testing equipment, I honestly have no idea how I would cope with daily life. The only real options for consumables are proprietary ones, which allow companies to compete in a price war between various proprietary solutions, monetarily playing the lives of T1D patients off against the state apparatus that keeps them alive.

The reality of the hardware that I use is that I am utterly dependent upon a proprietary framework that I am not allowed to understand, recreate, repair or deconstruct in order to administer my care. These issues mostly end up in economic territory — because companies create proprietary hardware, the results are unaffordable technologies that value the profit of private medical companies over the welfare and freedom of T1D patients who would die without the technology. I can’t just ‘install (GNU plus) Linux’ here, I am a captive audience for these proprietary technologies, with the alternative possibly being detrimental to my health. These are structural issues which, due to my privileged position, do not immediately threaten my livelihood (although I am vulnerable to changes in legislature), however I am deeply uncomfortable with corporations having an intellectual property clause over what is effectively one of my organs.

The proprietary software I use to manage my T1D on the other hand, is a more immediate issue for me.

I use an insulin pump to administer insulin. The way an insulin pump functions is to use an infusion set to administer a continuous stream of insulin. This is my sole method of insulin delivery (except for emergencies), and I administer doses by typing a dose to the pump that I use, which then administers the relevant dose. This is a wonderful system in terms of convenience — I can easily administer insulin doses as part of my daily routine without any break in whatever activity I am performing (the alternative being manually injecting myself after every meal). However, this pump uses proprietary software including a proprietary Bluetooth® remote which is also capable of fully operating the pump wirelessly. The problems begin here.

In the above Defcon talk, Jay Radcliffe (himself a Type 1 Diabetic) gives an account of penetration testing his own insulin pump, showing the possibility to turn off an insulin pump remotely. The implications of this demonstration are enormous — if you can turn off a pump, can you administer a dose remotely? The answer is (probably) yes.

Via The Register
Via Reuters

The proprietary software that these insulin pumps (and probably others that have not yet been investigated) run is vulnerable to hacking, and that hacking can have grave consequences. At any one time I can be carrying up to 315 units of Insulin in my pump reservoir, and if this was discharged without my knowledge it would send me into an acute hypoglycemic attack, and could well kill me. Because the software and hardware is proprietary however, we have really no way of knowing if this is a reasonable risk, and Bluetooth doesn’t have a stellar reputation for security. Proprietary software is often not properly audited as there is little to no accountability for writing bad code if nobody can see it, and proper security auditing costs money that does not result in an immediately observable outcome to consumers of proprietary products, so as a cost-cutting exercise it is simply not done. We have seen this over and over in many stories surrounding poor security (both in bugs in proprietary software as well as poor security practice) — Moonpig, Thomas Cook, Voting machines, Adobe Flash, a smart gun and Many ‘Internet of Things’ devices to give a few examples.

In terms of medical devices, Insulin Pumps are by no means alone in being vulnerable to hacking either, WIRED reports that medical devices are the next security nightmare, with a talk from Defcon 23showing a myriad of ways in which an individual could digitally (and physically) infiltrate a hospital and gain control of the equipment within it. All things considered I am not filled with confidence about the security of my own organs — sure it may be a theoretical risk, but it’s a risk I would rather not take. I am lucky that my pump gives me the option to turn off Bluetooth control and I have had this functionality on my pump turned off since I saw Radcliffe’s talk. Some users such as Radcliffe however, do not have this option, as the developers simply did not add one.

Another deleterious aspect of proprietary software that Stallman mentions in his talk is surveillance. In 2010, Apple updated its privacy policy to leverage its software to track the location of iPhones. With proprietary software, if you disagree with this aspect of Apple’s practices, there is no way to turn it off or remove this functionality — this ability is signed away when a user agrees to the Terms of Service. All digital pieces of equipment I use to manage my diabetes can be connected to a computer to download data, and the software that is used to download this data is proprietary. It is entirely possible that at any point the software used to download the data could have functionality added that would send my data to insurance agencies for example, and I would have to give up that software in order to preserve my privacy (I can’t run the software anyway as it is not available on Linux). I don’t hold out a lot of hope for the future either, I fear soon an insulin pump I use (I have to get a new one every four years due to warranty expiration) will be internet-connected by default, with software automatically updated over WiFi, which uploads my personal medical data to a company-controlled cloud which distributes my data covertly, in a way reminiscent of Apple, Facebook, Microsoft, Google, Unroll.me, or Roomba. The alternative to this? A direct, life-altering decline of the ease of my medical care in my medical care in exchange for some privacy.

All of this is quite grim. Do I have any other options?

I’d absolutely love for the software (and hardware) my body runs on to be as free as the software my computer and phone run on. I’d love to be able to make an ethical choice about the kind of software and hardware that governs my body, but I can’t. In his talk, Stallman says ‘there may be some inconveniences you have to suffer, for your freedoms’ sake’, and I have done that for every aspect of my computation. I don’t use Apple machines, I avoid using proprietary software to make art or perform general computational tasks, and I avoid proprietary services wherever possible — and I did have to suffer some inconveniences, including having to re-learn my entire computing experience. However, the alternative to proprietary software serving as one of my organs is currently to remove any digital alternatives at all — my only hope of treating my T1D in a way that does not limit my life too much. I am not comfortable with that.

There are some ways to adopt FLOSS software in my diabetes care, such as OpenAPS, however most FLOSS solutions are focused on the management of diabetes data such as blood glucose and insulin doses, such as Nightscout and Tidepool. The reason for this is that developing FLOSS solutions to T1D management that includes insulin delivery has some very obvious dangers. My Linux laptop is not the most reliable, and sometimes will crash, or get stuck in a loop which maxes out a core of my CPU, or does some unexpected things due to graphics glitches — this is not a problem for me as I can simply reboot the laptop or dive in to the terminal to turn off individual processes. With my body however, there is no space for error. If I mis-dose my insulin as a result of a software bug, the results are potentially catastrophic, or at the very least a real bodily inconvenience. I can imagine this is a large reason for the lack of FLOSS medical equipment, as generally FLOSS wares are distributed without any kind of warranty or liability on behalf of the copyright owner. Take this example from the GNU GPL 3.0:

  1. Disclaimer of Warranty.
THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.

My ideal solution to T1D care would be a 3D printed pump with a model under a copyleft license, using free hardware with FLOSS firmware, with low-cost generic (in the sense of Generic Drugs) or 3D printable consumables. This could be linked up with a similar system for Interstitial Blood Glucose monitoring, relying on generic sensors running FLOSS firmware working using open standards for data transmission which communicate with a 3d-printable blood glucose meter running on open hardware and software. Any transmission of data which could result on an insulin dose should be cryptographically secured, and protected from hacking — this can be verified by users of the code, as well as by independent security auditors, and this process would be conducted in the public domain.

I am aware that these hopes for my care are very optimistic, and that is okay. I will probably never see an ethical, affordable, readily accessible FLOSS solution to my care in my lifetime unless there are dramatic changes in legislature and market forces. And, short of making these things myself, I am not really qualified to speculate about, let alone contribute to, this vision.

That said, this is something that affects me every minute of every day. I am dependent upon the machinery of corporate capitalism every day to simply not die, a fundamental homeostatic process that my life depends on is subject to warranties, fair use clauses and copyright. In an era of late-capitalist anxiety for people of my generation I’ve often heard my peers fantasise about living ‘off the grid’, but not only will I die without the grid, The legal framework of capitalism is physically embedded in my body as I type this. I am beholden to companies willing to extort diabetics with their lives, and provide frankly shocking services for the products they provide because they have a physically dependent audience (I have some anecdotes about dealing with Abbott Laboratories). All this in addition to having to deal with the drudgery of having an ‘invisible disability’, a chronic condition that follows my every move, where a lack of constant attention can result in anything from irritation to amputation. An epidemiological study in 2008 revealed that ‘The prevalence of anxiety and depression symptoms in patients with diabetes is considerably higher than in general population samples.’, and after writing this, I’m honestly not surprised.

I’ve been chewing on these issues for a long time, and how to extricate myself from this difficult set of relationships between body, chronic conditions, copyright and capital. My current thinking is that in absence of a concrete bodily solution to these problems, a decent solution would be to tackle some of it through art or performace, even if it only serves to raise awareness or disseminate information about this nexus of issues. I’ve never dealt with such personal issues in my work as an artist before, but I feel as if the time has come to tackle this, and make something entertaining, informative, beautiful, and probably a bit difficult.

I’ve tentatively titled the project ‘howToNotDie’, and as I always do I will be updating it on GitHub as it progresses. I’m hoping to produce a creative response to these realities through free software, of which this is the first part.