At this point in the evolution of healthcare data, you’d think it would be easy to at least define interoperability, even if we can’t make it happen. But the truth is that despite the critical importance of the term, we still aren’t as clear as we should be on how to define it. In fact, the range of possible solutions that can be called “interoperable” is all over the map.
For example, a TechTarget site defines interoperability as “the ability of a system or a product to work with other systems or products without special effort on the part of the customer.” When defined down to its most basic elements, even passive methods of pushing data from one to another count is interoperability, even if that data doesn’t get used in clinical care.
Meanwhile, an analysis by research firm KLAS breaks interoperability down into four levels of usefulness, ranked from information being available, to providers having the ability to locate records, to the availability of clinical view to this data having an impact on patient care.
According to a recent survey by the firm, 20% of respondents had access to patient information, 13% could easily locate the data, 8% could access the data via a clinical view and just 6% had interoperable data in hand that could impact patient care.
Clearly, there’s a big gap between these two definitions, and that’s a problem. Why? Because the way we define baseline interoperability will have concrete consequences on how data is organized, transmitted and stored. So I’d argue that until we have a better idea of what true, full interoperability looks like, maybe we should map out interoperability “bundles” that suit a given clinical situation.
A Variety of Interoperabilities
For example, if you’re an ACO addressing population health issues, it would make sense to define a specific level of interoperability needed to support patient self-management and behavioral change. And that would include not only sharing between EMR databases, but also remote monitoring information and even fitness tracking data. After all, there is little value to trying to, say, address chronic health concerns without addressing some data collected outside of clinic or hospital.
On the other hand, when caring for a nursing home-bound patient, coordination of care across hospitals, rehab centers, nurses, pharmacists and other caregivers is vital. So full-fledged interoperability in this setting must be effective horizontally, i.e. between institutions. Without a richly-detailed history of care, it can be quite difficult to help a dependent patient with a low level of physical or mental functioning effectively. (For more background on nursing home data sharing click here.)
Then, consider the case of a healthy married couple with two healthy children. Getting together the right data on these patients may simply be a matter of seeing to it that urgent care visit data is shared with a primary care physician, and that the occasional specialist is looped in as needed. To serve this population, in other words, you don’t need too many bells and whistles interoperability-wise.
Of course, it would be great if we could throw the floodgates open and share data with everyone everywhere the way, say, cellular networks do already. But given that such in event won’t happen anytime in the near future, it probably makes sense to limit our expectations and build some data sharing models that work today.