One of the first things a new developer in healthcare quickly realizes is integrating and launching your application on top of medical data can be a total nightmare and often a showstopper. There are hundreds of electronic medical record (EMR) vendors and every implementation has a different “flavor” of a handful of competing standards.
That is all about to change. There is a growing movement emerging in Health IT to finally begin standardizing the way applications can access data. The preferred method is through what is called an application programming interface, or API.
APIs have been notoriously absent in health IT until very recently. This year has seen a surge of enthusiasm and support for APIs specifically around facilitating access to EMR systems. Some of these initiatives have been driven by the EMR vendors themselves, and others are cross-vendor collaborations such as HL7 FHIR.
When Investment Rhymes with Canada
Canada has a proud history of achievement in the areas of science and technology, and the field of biomanufacturing and life sciences is no exception.
APIs in general originally emerged in software as a great way for companies and vendors to share the functionality they’ve built in their proprietary application with a wider set of developers.
The key feature of an API is that it allows a developer to treat the application like a black box. You send a request, let’s say for a piece of data, and then the black box answers with the requested piece of data. During this process you don’t see how this action is performed by the application, only the result, which preserves the proprietary source code of the application while still letting you leverage its functions.
The concept of an API is deeper than at first glance. By allowing you to ignore the “how” of a particular application or solution, you are effectively achieving a new level of abstraction. The concept of abstraction is a powerful one in programming. As the complexity of developing applications continues to grow, it’s crucial to be able to remove the details of certain components and reduce them to the bare-minimum needed to use (and hopefully reuse) them. This is one of the core drivers of object oriented programming and other techniques for efficient development and architecture creation.
What’s so special about APIs is now you’re not just abstracting a piece of code, but an entire application. This allows developers to build their code on the shoulders of whole other teams of developers who’ve focused entirely on solving a specific set of problems with their application. In many cases this drastically speeds up development.
This acceleration factor has made APIs extremely popular across all major verticals. Some of the most famous examples are Facebook’s API, which allows developers to incorporate Facebook functionality and Google’s YouTube API which let’s programmers integrate YouTube videos into their applications or websites.
A powerful example of what’s possible with APIs is the recent release of TensorFlow by Google. In March 2016 Google recently made headlines because AlphaGo, their artificial intelligence (AI) program, beat the reigning world Go champion 4-1. Until that point beating a human champion at this particular board game was considered one of the more difficult open problems in modern artificial intelligence research. This is mainly due to the difficultly of simply “out computing” the opponent (a standard 19-by-19 Go board configuration has 10170 legal board configurations. For reference, this is roughly double the 1080 atoms in the observable universe. )
After many years of effort and millions of dollars of R&D, Google’s DeepMind team figured it out. What happened next was game-changing: instead of keeping the whole project to themselves they released major portions of the code and framework as a set of open APIs and libraries called TensorFlow. This instantly democratized AI programming for millions of developers. Now anyone who wants to build an application using cutting edge AI tech can leverage the TensorFlow API. With a few lines of code you can skip decades of AI research and go straight to the cutting edge of the field.
Another real-world use-case would be weather prediction. Unless you also own a ton of sensors, satellites and a vast array of other global infrastructure it’s unlikely you can predict global weather patterns on your own. This means that if your application wants to incorporate current weather or forecasts it has to rely on a third-party weather company, like Wunderground.com. Without APIs the only way to get the weather forecast would be to input the human-readable website and analyze it for patterns that will let you extract the needed information. This is a ton of work and worse it risks breaking the moment even a small update is made to the website. Luckily Wunderground.com has an API your app can call to get the weather from any location in a format that is easy to use and consistently predicable. Now having a current weather feature in an application is trivial.
Ok, so we’ve established that APIs are great tools. Why do they matter so much for healthcare? Two reasons: Time and cost.
The traditional way of integrating into various EMRs required an application developer to recreate a new data parser, integration engine and data mapping schema for every single site they wanted to deploy to. What this meant was spending hours of development time re-inventing the integration wheel instead of focusing on taking the data stuck in the EMR and transforming it into valuable applications. This extra work translated into extra costs to bring an application to market, meaning that the entry cost to innovate with healthcare data was extremely high. What this work also meant was a slower time to market even if you could get past the technical hurdles.
What these two challenges have translated to is a healthcare system that is chronically on the trailing edge of technological innovation. And in this industry that can mean potentially life-saving technologies are not being deployed or even built because of the complexity and cost of getting them off the ground.
All that is now changing. As APIs become available for accessing and handling the data previously stuck and siloed in EMRs, the cost and complexity of launching has been drastically reduced, finally clearing the way for innovative companies to build data-driven solutions to a wide variety of healthcare challenges. What this means is we’re going to witness a surge in the number, scope and ambition of healthcare applications, and with it, hopefully, an equally drastic improvement in the quality of care for all of us.
Isaac Newton famously said “If I have seen further, it is by standing on the shoulders of giants.” Now, with a little help from APIs, so can you.
Photo: Flickr user Lóránt Szabó
