Let me let you in on a little secret: There is no meaningful interoperability of patient healthcare data right now.
Why? We can choose to point fingers at electronic health record (EHR) vendors for not caring enough or government and standards committees for mistakenly thinking interoperability can be regulated into existence. But in the interest of being productive, let’s take a step back and look first at what we mean by “interoperability” and whether there is a market that can support this idea.
Why Is Interoperability So Difficult to Achieve?
In a sense, we’ve had interoperability for 30 years — HL7 standards were first established in 1987 and IT systems in hospitals share messages about patients all day, every day. However, what people really want from interoperability is for patient health data to be easily shared across different health systems, geographic areas and health facility types. This version of interoperability is the only kind that will deliver better, more efficient care for patients and improved health outcomes at lower costs.
However, “real” interoperability is not and cannot be delivered with the approach we have taken so far. The problem of systems interoperability is not technically difficult. A handful of strong engineers from various industry stakeholders could make short work of it. Frankly, we’ve seen other industries accomplish more difficult interoperability feats. For example, connecting all the nation’s banks into a financial system, shipping goods worldwide and the connection of our computers on the internet are all equally hard, if not harder, problems that have been solved.
So what’s the issue?
Set featured imageSimply, health record vendors don’t have any real incentives to easily share information across systems – in today’s market it gives them no competitive advantage and may actually hurt them.
First, Take a Look at the Market Landscape
There is currently no compelling business need for interoperability.
This is not the same as saying that patients and healthcare systems wouldn’t benefit from interoperability. Few would argue that a comprehensive, longitudinal patient health profile wouldn’t lead to better outcomes, greater insight into care and better control over expenditures. However, this is a theory and not a practice — in product development parlance, interoperability lacks strong use cases.
Broad interoperability initiatives are driven by the notion that interoperability is an intrinsic good. They are often centrally controlled and expect participants in the market to ante up so that everyone can benefit. It’s a classic “field of dreams” scenario. In the current state of affairs, those chasing interoperability are trying to build a product that has no defined target user and doesn’t clearly fit into any existing workflows. Regulating a standard into existence does not solve a well-defined problem, which is largely why current standards haven’t led to acceptable interoperability results.
We know this is true experientially, by looking at a sector that many of us are more familiar with: social media. My data in Facebook is not interoperable with my data in Google or LinkedIn. I can’t see my friends’ Facebook updates in my Twitter feed nor can I see an acquaintance’s career history on Facebook. While combining all of these updates into one feed may save users’ time, there is ultimately no incentive for these social media companies to execute this type of exchange. Furthermore, nobody is demanding it or threatening to discontinue their use of the service. With the absence of these incentives, it won’t happen.
However, both Google and Facebook would like more content posted to their networks, an example of a use case, so as a result various apps we use have the capability to cross post to multiple social sites. (We see this all the time with medical records, but it’s usually “cross posting” via a fax machine and scanner.)
Take a Page from the Music Industry’s Playbook
Instead, the best strategy is to start at the grassroots level and move up. Take one issue we are trying to solve and build a targeted solution for it. It’s not glamorous and it’s not a silver bullet, but it might actually work.
For example, take the way that the music industry reached interoperability for digital music. In the 1980s when synthesizers first emerged, they were not interoperable with each other. Since each manufacturer’s keyboard made a distinct set of sounds, musicians grew accustomed to using layers of keyboards to develop complex sounds drawn from the sonic capabilities of many companies.
Companies could see that lack of compatibility was hurting sales, so in 1982 the leading American and Japanese synthesizer manufacturers came together and agreed on a set of universal standards for digital communication, the MIDI standard. MIDI was efficient, simple and customer-centric, and soon adoption took off — not just in synthesizers, but across the music electronics space — creating the digital music industry as we know it.
Old school technology? Without a doubt the technology challenge in healthcare is far greater, but the key here is that the manufacturers themselves developed a standard in response to a specific need and then made it work and scale. Market forces pushed competitors to collaborate and create the foundation on which interoperability was possible.
Understanding the True Goals of Interoperability
Do we want to allow doctors treating the same cancer patient to share info about checkups, medicine updates and lab and radiology tests? Do we want researchers to have quick access to relevant clinical data? Do healthcare providers need a way to pull data for analytics and reporting? Is there a need for securely transferring patient documents between two EHR systems in real time?
If so, then let’s devise a system that shares the necessary information between the relevant stakeholders so they can do their jobs better. And guess what? If the solution provides value — whether it’s improving productivity, lowering costs, boosting health outcomes or improving patient care — then it will scale organically.
One of the main health interoperability coalitions, the Commonwell Alliance, seems to be moving in this direction; it lists a use case for acute care and a use case for ambulatory care on its website. And startups like Flatiron Health are building business models for interoperability around very specific needs (in Flatiron’s case, cancer care coordination). Yet we need to move further and faster in this direction.
It’s time to let go of the build-it-and-they-will-come approach. To be successful, the industry must stop viewing interoperability as an inherent good that will inevitably confer benefits if everyone would just play the game fairly. Rather than continuing to fall down the rabbit hole of current interoperability practices, let’s define what the benefits of interoperability are, one concrete use case at a time and pursue these as market opportunities.