The ginormous spring HIMSS (Health Information and Management Systems Society) meeting is ending. The big big subject this year? Interoperability — the ability of different systems to link up, transfer data, and actually operate through each other. Here’s why it’s a big deal:
The lack of real transparency and interoperability across the larger systems we are building, across accountable care organizations, even into other institutions that our customers may go to, often even within our own organizations, the failure to produce truly clinician friendly customizable interfaces in most of healthcare IT, the failure to create open systems that can work directly with other pieces of software — all of these together are a patient safety and quality disaster. When a clinician cannot get the real information on a patient in a way that is instantaneous, accurate, accountable, in an interface that helps them to insight rather than getting in their way, patients suffer.
Epic, the dominant force in the market, claims its systems are interoperable, but that is true only for very special use cases, and for a limited definition of “interoperable.” This week Epic announced a new open systems approach, but it was merely talking about APIs through which other vendors can build software modules that can connect into the Epic mothership, not about connecting to other EHR systems across town. In the meantime some organizations are joining the CommonWell Health Alliance to bring together non-Epic data standards. Many providers are dragging their feet from a combination of IT fatigue, stretched budgets and a “What’s in it for me?” attitude. The Office of the National Coordinator (ONC), the federal agency in charge of trying to whip all this into shape, has set a goal of true seamless healthcare data exchange real soon now. Well, by 2025. And it considers that an ambitious goal.
Quite simply this situation is killing people. But the disaster goes beyond that. Lack of interoperability makes the vision of truly seamless care across accountable care organizations, let alone regions, impossible. The vision of the Next Healthcare beyond the tipping point is built on a seamless data universe centered on the patient and the patient’s family and caregivers, not on the hospital billing department. Not getting the information flow right will make the hope of better, cheaper healthcare much harder to reach.
The bars to making this work are mostly not technical. Building these vast information systems is very hard. Making them open and interoperable is not the hard part — if the will is there in the vendors, the providers, and the payers to do it.
it’s 2015, and it is at last dawning on the health care “industry” that exchanging information is an essential feature of a viable health care system.
It would be smirk-worthy, if it weren’t so pathetic, so deplorable, so borderline sociopathic.
I am convinced that the payer does not want to know best practice. If we were the insurance company we would surround ourselves with the most accomplished clinical researchers from the best Universities and we would ask them to tell us who we should pay how much for what. We tried for years to gain access to their database without success. We have the mechanism to analyze their data to determine effectiveness and efficiencies correlating intervention to spending.
I continue to be concerned about the ultimate utility of the current darling HL7 FHIR API thrust. I fear we may end up simply making a bigger mess. Any one clinical datum is essentially useless by itself. Context is everything, and the breadth of the data comprising that context is often quite large for any one patient, going beyond even the gamut of data collected during any one encounter (e.g., flowsheets, trends more broadly).
Will we end up with scattershot atomistic data snippets floating around in recipient EHRs? Will we simply define down the IEEE definition of “interoperability” (“…without special effort…”) so that seamless, comprehensive, ongoing data exchange does really not happen? Will these “document-centric” XML-oriented “data-a la carte” API structures end up being faxes by any other name? Yet another workflow-addition “in-box” to manage?
I hope I’m wrong. I didn’t go to HIMSS15 (beset by my own health issues of late), or I would have been pressing for specific, multi “use case” answers.
http://regionalextensioncenter.blogspot.com/2014/10/interoperability-solution-hl7-fhir-we.html