The Last Mile-Data Integration
There is no doubt about it, we are surrounded by innovation and technological advancement. I can’t remember the last time that I have gone a full day without feeling awed or delighted by some new invention. We are finally at a point where scalable computing power is ubiquitous and inexpensive. This has really opened the doors to implementing great ideas that would have been non-starters just a few years ago.
So why does it feel like it takes so long for these amazing innovations to find their way into healthcare? In my opinion, there are a few major contributors (i.e., limited reusability, individual BAAs,), but the main issue is a lack of integration. I worked in healthcare information technology for years at the Cleveland Clinic and evaluated technology solutions. Time and time again, I saw elegant solutions to long-standing problems fall to the wayside because there was no integration. Healthcare IT is an environment where clinical users count clicks and seconds. In this setting, the idea of switching from the electronic health record (which MUST be used) and logging in to a separate system is enough to spell doom for even the best technology. Failing this test is the quickest path to rejection.
As important as this level of integration is to clinical workflows, healthcare software developers frequently skip this part of the project. Historically, integration has been costly, time-consuming, and site-specific. As such, it is the first feature to be abandoned when budgets get tight.
A couple more subtle (but vitally important) considerations are data quality and data availability. Health information is generated by multitudes of people and managed by a variety of clinical systems. Diametrically opposed to this architecture sits the ideal, in which clinical decisions are evaluated with the most current and highest quality information. This ideal enables clinicians to have full access to the information they without investing extra time searching multiple systems. The electronic health record must be not context-sensitive, but comprehensive and trustworthy.
This is the last mile – connecting external systems and their data with the health record, using a single sign-on solution. The first 999 miles mean nothing if the last mile is ignored. SMART on FHIR provides the necessary boost to get through the final mile. Based on open standards (OAuth) and familiar message formats (JSON), SMART on FHIR is positioned for widespread adoption. It is lowering the barriers to entry dramatically, leaning on well-documented specifications and a burgeoning set of development tools.
I am tired of great solutions failing before they get a chance to shine. Let’s get them plugged into clinical workflows and start shifting things toward a value-based reality.
This article was written by Mike Speck, a founding partner of nuboHEALTH and serves as the President. Mike provides leadership in positioning the company at the forefront of the healthcare technology industry and sets the direction for future technology development.