Digital healthcare: Should doctors learn to code?
The title of this blog suggests what many would consider as unreasonable. Doctors already spend years trying to become an expert at treating patients. I should know better than to suggest making a grueling career path even more harder because I went through it myself, so I would request you to hear me out before coming to that judgement.
I was another teen of the 80s. Yes I loved the outdoors but there was also a reasonable dose of computing available at home. I was fortunate to have with me the Atari game console and the coveted home computer, the precursor of today’s PC. Many of you from those years will remember the Spectrum computers and how its 128KB was a lot of memory compared to the very few options out there for home use. I spend days and nights playing the familiar PacMan, space invaders and every other game I could get my hands on. Those games did not just provide entertainment but the creative beauty in them simply fascinated me. The out of the box programming language on these computers- BASIC – was too rudimentary that I imagined the game developers had access to something more powerful. I believed that once I got a hold of this arcane knowledge I could do incredible things. I was so excited about the possibilities that I decided to go to the core of programming. I put the games away and went on an inspiring journey learning assembly language, a primitive coding tool, which to put it simply allows you to talk directly with the CPU. Learning machine code turned to be one of the most exciting things I had ever done. I began to ‘see’ what goes on inside these machines and how an interplay of logic and numbers could result in all those addictive games. All through my clinical career and now in my work in the digital arena, I remain fascinated by the fact that numbers and code can create so many visceral experiences.
I was reminded recently of this power of software when I read an article in MIT Sloan Management Review where the authors emphasize a neglected truth -algorithms and the code are not value neutral. The developer’s values, aspirations and their beliefs of what humans should be are hard-coded into software. The authors explain this with the example of the programming behind self driving cars. What is the algorithm, they ask, programmed to do in a situation when the only choice is between hitting a pedestrian or an action that will result in the death of the driver? These are very important questions as man indulges further into artificial intelligence. Healthcare is another industry on that journey albeit moving at much slower pace.
Before we even get to AI let us consider the burgeoning amount of knowledge that is being generated in research of every kind and especially medicine. While computers and software have embedded themselves comfortably inside healthcare, we are surely on the cusp of a digital revolution where digitizing this increasing amount of knowledge will make medicine more manageable and thereby delivered with greater perfection. However, we must not forget that revolutions cannot be leaderless. There are followers, people and code, waiting to be led and nobody in healthcare is better suited for this role than the providers themselves. They can lead by ensuring that medical knowledge is transferred, in the most appropriate way, into digital processes and algorithms . Whether the end result are mobile apps or powerful machine learning algorithms the code beneath these artifacts is not value neutral and the best people to make choices would be doctors who also understand how machines work. Doctors with this kind of dual expertise will be very effective, when working with software developers and data scientists, in ensuring that healthcare maximally leverages the power of digital capabilities. Here I will also make a suggestion contrary to where this blog seemed to have stared out. I am not really suggesting that doctors sit and code but instead use their understanding of how code works to lead developer teams as well as to manage the algorithms and digital processes that are the outcome of digital initiatives. Technology vendors (including the one I work for) on the other hand have a responsibility to make available software development platforms in which this kind of cross collaboration and co-development are key objectives and not an after-thought. Fortunately, technology today makes it possible to do this to a great degree. I would even go as far as to say that technology is waiting for healthcare to take advantage of it.
Image Credits: http://www.gondolin.org.uk/hchof/machines/spectrum