Back in the late 90s, I read an article in a trade magazine and the piece made such an impact on me that the ideas articulated have stayed with me to this day. The article, “The Rise of the Stupid Network,” was written by a man named David Isenberg who worked at AT&T Labs Research.
In the article, Isenberg describes how telephone companies of the day were basing their infrastructure decisions around key premises, including the following:
- That expensive, scarce infrastructure can be shared to offer premium services
- That the communications infrastructure of the day (at that point, primarily circuit-switched calls) was the only communications technology that mattered and
- That the telephone company is in control of its network
In making these assumptions, Isenberg wrote, the telcos were tuning out several important trends, including:
- The massive annual growth in the volume of data traffic running on the network
- The many different types of data that had begun to travel on the network
- The diversity of communications technologies under use that were not part of the core network and
- The extent to which the Internet was shifting control to the end-user.
In his view, telcos like his employer should not have been blindly working to integrate features like more automatic operation and intelligent new services into the existing network architecture. Despite their efforts, he argued, their Intelligent Network was rapidly being superseded by a Stupid Network offering nothing but dumb transport in the middle and intelligent user-controlled endpoints.
While the analogy is far from perfect, many ideas here could be relevant to the EHRs of today. What I see happening is a shift to putting the most critical functions providers use at the edge of its network or even in completely independent hands.
As with the Intelligent Networks of the 90s, EHRs are being designed and built with the assumption that using powerful centralized technology they will manage the data their customers used, and that if customers had any other needs the relevant technologies would merely be tacked onto the core EHR infrastructure.
Yeah, I hear you saying, so what? It’s not abnormal for tech vendors to build a ring of related applications around it that integrate with its core platform and share data with end-users. In this scenario, the EHR is still the hub of all data activity and decidedly the controlling partner in the mix.
Still, EHR vendors are beginning to recognize that some of the most important capabilities they need to deliver can’t be put into place casually on their infrastructure.
Among the most recent examples of this are the deals both Epic and Cerner have recently made with digital health vendor Xealth, which offers centralized digital ordering and management of digital tools on an independent platform.
What we really need, however, is to move beyond systems that create point-to-point, EHR-to-EHR connections between data sources to creating a web of health information. These sources will include a much more robust system making sense of data from well-equipped smartphones, sophisticated, medical-grade wearables, tablets, voice assistant data and far, far more.
And while the EHRs housing this data today will likely never become completely obsolete, the job may increasingly become a “Stupid EHR” whose job is more to switch information from user to user in much the same way telephony systems do.
The truth is, thinking of EHRs as virtual versions of the paper and file cabinets providers once used is already out of the question. What we haven’t decided, but need to soon, is how far the critical functionality supporting digital health must be pushed to the edge of the network.
My feeling is that the trends relegating EHRs into just a part of a data universe are unlikely to move quickly until the pandemic truly begins to loosen its grip. When that happens, though, it could be that today’s EHRs will have a far different role to play in data sharing’s future.