Open source is the only way for medicine

Software is now medicine itself—and proprietary code is putting profits before patients.

Medicine affects human life so profoundly that we need to consider the morality of medical technology differently to other industries. In other tech areas the consumer usually has more choice – including choosing not to use that technology. But people can’t ‘choose’ not to be ill, or not to need healthcare.

When people are denied access to the best available treatment, the consequences are severe: unnecessary pain, avoidable disability, or even early death. Unlike missing a game release or a Windows upgrade, medical care is genuinely life and death.

This is why I think that closed, proprietary medical software is unethical. As medicine becomes reliant on software to deliver healthcare, this closed-source and proprietary medical software undermines our ability to share medical knowledge, and puts intellectual property rights ahead of human rights. Open source is the cure.

Software has one unique advantage over hardware or physical technologies: it can be copied infinitely for almost zero cost. So, it’s not a zero-sum game – when I share my open code, I’m not left with any less code myself. So if we have better software that can help us deliver better care, why wouldn’t we want that software shared as openly as possible?

Software IS medicine now

My case would have been weaker, I guess, a few decades ago. Back then, medical software was very basic: systems for patient administration, databases for medical records. Nothing particularly exciting, and nothing likely to make a big difference to patients’ lives. These dull admin systems weren’t considered to be ‘part of medicine’, they were like a card index.

Now it’s very different – from closed-loop insulin pumps for diabetes care, through genomic analysis tools, to algorithms for predicting illness, complex software is now integral to medicine – a good system will deliver better care (and a bad system could cause harm!). Yet, in most cases medical software is closed-source, proprietary, non-peer-reviewed, and not testable.

As software becomes more and more part of medicine itself, we have to insist that this software becomes part of the shared body of medical knowledge, and open-sourced, for maximum benefit to humankind.

Read more: Why open source is critical for the continued advancement of new tech

The copyright time bomb

We’re in a world in which software is a medical treatment, like a drug. But here’s the catch that nobody talks about: The difference between patents and copyright.

With drugs, the pharmaceutical company files a patent for a new drug, which allows them to recoup their R&D costs, but then it expires after about 20 years, enabling the drug to be produced ‘generically’, at a much lower cost. This step drives down the costs of established drugs, and is vital for affordable healthcare. It also forces drug companies to innovate and come up with new drugs.

But with software, we’re dealing with copyright, not patents. Copyright lasts far longer – typically 70 years or more – meaning that medical software algorithms copyrighted today now will never become ‘generic’ in your lifetime. Somehow we’ve sleepwalked into a worse situation with software than we were in with pharma!

As algorithms become more central to medicine, we’re building up a legacy of proprietary medicine, which to all intents and purposes will never become ‘public domain’. And I’m not talking about the code here – nobody is going to want anyone’s 70-year-old code, right? Intellectual property rights cover the underlying algorithms, too. 

Global healthcare just can’t afford for every treatment to remain proprietary and full price, forever. Also, when they can ride us for ‘full proprietary price forever’, what is their incentive to innovate?

There is a defence against this scenario. You guessed it: open source.

Medicine’s open source tradition

Medicine has historically always been an ‘open source’ profession, even before the term open source came into being. We have always had a culture of sharing knowledge and passing it down to the next generation of doctors that we train:

The Declaration of Geneva, the modern successor to the Hippocratic Oath, makes this knowledge–sharing responsibility very clear:

 ‘I will share my medical knowledge for the benefit of the patient and the advancement of healthcare’. — Declaration of Geneva

We’ve always had this duty to share and teach medical knowledge. But the gradual creep of proprietary software threatens our ability to do so, since medical knowledge is increasingly built inside proprietary software.

The cost of secrecy: A historical warning

History gives us a scary demonstration of the human cost of ‘secret’ medical treatments: the Chamberlen Forceps. These early obstetric (baby delivering) forceps were kept secret by their inventors for around 150 years, The Chamberlens used their proprietary technology for Europe’s rich and royal families only, even though the idea could easily have been shared and the forceps were easy to make. During that 150 years, we’ll never know how many mothers and babies’ lives could have been saved by forceps delivery.

The ‘secret’ Chamberlen forceps were discovered by accident centuries later, revealing how the Chamberlens had been delivering babies, Because of such precedents, modern medical treatments have to be openly published in respectable medical journals before they are taken seriously. It would be unthinkable now in the medical profession to try to offer a ‘secret’ treatment. But for medical software, it seems to get a free pass.

Verification and transparency matter

As medical software becomes more life-critical, we have to be allowed to see inside it, to verify that it really does what it says it does. We need to be able to prove it works, which means external code review, safety appraisal, and open bug tracking. Do we have any of those things? No, of course we don’t, because the software is proprietary. As clinicians, we have to be more careful criticising healthcare software publicly (for example pointing out bugs) than we do criticising pharma, because we have no protection against software company lawsuits.

The profession of medicine has had ample opportunity over the centuries to see the dangers of quackery, alchemy, and witchcraft – which is what happens when developments in medicine can’t be independently and scientifically scrutinised.

Open publication of medical research is completely accepted as normal. But the software we use in medicine seems to get a free pass to be closed.

The path forward

The medical community should actively take steps to prevent the further ‘proprietorization’ of medicine, and going forward we should insist on open source software for all of medicine.

It’s an uphill struggle though, because many in healthcare technology are ignorant of, or naively opposed to, open source. The healthcare software industry is heavily incentivised to maximise profits, and (possibly related?) our UK health service recently removed all its open source policies without explanation.

Education is key! Start by asking questions: Is the medical software treating you or your loved ones open source? Can it be independently verified? Healthcare providers and policymakers must hear these questions repeatedly until open source becomes the standard, not the exception.

Acknowledgement

I’d like to thank Tony Shannon, Principal Officer at the Office of the Government Chief Information Officer in Ireland, for first making me aware of the Chamberlen Forceps story.

More from We Love Open Source

• 12 everyday technologies powered by Linux
• 10 open source tools you can start using today
• 4 ways your company can support open source right now
• Why open source is critical for the continued advancement of new tech
• Agent Definition Language: The open standard AI agents have been missing