Open Source Health

When I think of open source and healthcare, I always think of VistA from the VA. I’m not sure how viable VistA is as a production EMR, but Oroville Hospital certainly got a good amount of press when it announced it had customized and rolled out an EMR based on VistA.

I also think of OpenMRS, which is extremely popular internationally and in public health, with a very active open source development community. I’m still a bit surprised that nobody has written any systems or apps in the US based on OpenMRS. I’m not quite as thrilled with the UI of OpenMRS, but the back end seems pretty powerful, especially for population health stuff. Are there US-based software tools developed using OpenMRS? I’d really be interested to see them.

I’ve been thinking a lot about open source licensing and technology recently. In the process, I’ve been researching open source tools in healthcare, something that has taken me beyond VistA and OpenMRS. There are some really great open source stuff out there for developers of health and wellness. Some of the more interesting things on GitHub are under Children’s Hospital of Boston. It’s where all the Indivo and SMART Program code resides. If you haven’t taken a look at it then you should — it’s interesting to see what’s out there and freely available. There’s other Blue Button code and a bunch of EMR platforms and modules on GitHub as well.

The interesting part is that, despite these tools being available and ready to use for commercial or any other use (licenses vary but the ones I’ve seen for SMART and Blue Button are pretty flexible), none of the open source tools I find on Git are that popular — judging by the number of stars and forks — compared to other open source tools I come across. I imagine most health app developers are just building from scratch or don’t want to bother integrating larger open source libraries with tons of functionality the developer won’t be using, or doesn’t think they’ll be using or needing.

Personally, I really like that big non-health tech companies are actively building open source tools. Some of the best examples are Google (tons of open source developer tools like Angular.js and Yeoman), Twitter (does anybody build websites without Bootstrap any more?), and Learnboost (if you develop with Node, you likely use tools created and open sourced by Learnboost). The main open source example I can think of in health, and this is sort of a stretch, is Font Awesome (if you go to the site you’ll likely recognize the icons from apps you use everyday). The guy behind Font Awesome is product lead at Kyruus, a health tech company (big physician data), so Font Awesome has been adding more and more health specific icons.

The extreme example of open source for companies are those companies that open source everything and typically try to make money on enterprise support and services. The most recent example is Meteor, which has raised a bunch of money towards its mission of making it really, really easy for developers to build apps using 100 percent JavaScript. The Meteor library is freely available to download and use, though it’s still probably a bit early to use it for anything meant for production. I’m not sure how this model would work in healthcare and don’t think any open source EMR models have had much success with this approach. The healthcare market seems more than big enough for something like this, but I guess the healthcare enterprise may not be ready today or may not ever be the right target, especially if it isn’t building many of its own apps.

Open source can be a polarizing subject for discussion. I’m a big believer in the power of open source, in using open source tools in development, and in contributing back to the community in the process. I don’t think every line of code that companies write is differentiating for them or is intellectual property (IP), and sharing that code back with the open source community, either by allowing individuals to share it or to open source it as a company, is a powerful thing; I also think it helps with recruiting developers if that’s a sell for you.

To some extent, it comes down to what you want to compete on. As an example, I don’t think competing or creating barriers to entry on interfaces is sustainable in the long term, though there are some billionaires out there that may not agree with me or may not be too overly concerned with the long term. As another example, I also don’t like competing on a data model. I’d rather see companies and products compete on how they add value to the data, such as through analytics or value add services on top of the data.

The Health Design Challenge is a  great move to leverage crowd-design to create open source tools. If you haven’t looked at the challenge, it’s worth it to see some of the winning designs. Entries were just images, essentially mockups, but the ONC will be converting those designs into working code (presumably HTML, CSS, and JS) that anybody can use as a starting point or ending point for display of CCD data. As the challenge page states, "This will be like Bootstrap for health records…", and it’s true, at least for personal health records.

For companies and investors, it comes down to protecting IP and minimizing risk. I think historically, and I’d vote many enterprise companies (especially in health) are operating in a historical tech context, most companies protected themselves and didn’t release any code as open source. This has been changing with younger companies and generations, mainly outside of health and I think mainly in the consumer tech space. Also, Twitter’s new IP policy for employee inventions is particularly interesting. I personally hope we start to see more of a change in healthcare as well.

What do you think of open source tools in health and the role companies should play in it?

Travis Good is an MD/MBA involved with health IT startups. More about me.

  • Lynn

    While there are certainly passionate advocates of open source software, when patient’s lives depend on consistency and reliability, not to mention systematic identification and correction of bugs, and standardized development and testing processes, I think I would rather go with a product with a formal support organization behind it–and even then, all is not foolproof. While Vista is often held up as a model for open source and interoperability, keep in mind that there are 130+ instances of Vista used within the VA, and they are not as “interoperable” as many suggest. UNIX/LINUX became a mainstream product when Red Hat realized the opportunity for developing a formal support and standards organization, which has yet to happen with health care open source products–although the Open Health Tools organization (http://www.openhealthtools.org/) is certainly an initiative worth a closer look.

  • Travis Good

    Great points Lynn. Open Health Tools is a great initiative. I actually met and spoke with Rob Kolodner about it at HIMSS. I’m not sure how I forgot to talk about it in my post. Thanks for bring it up.

↑ Back to top

Founding Sponsors

Platinum Sponsors