Playing with FHIR: Health Data Interoperability’s Next Chapter
Like explorers searching for the Northwest Passage, the U.S. healthcare industry has been searching for the key to unlock health data interoperability. Thwarting the quest are ice floes of governmental bureaucracy; free-market healthcare economics rewarding providers and EHR vendors who keep data proprietary; and legacy IT infrastructure slowly transforming itself into a more modern, cloud-based model.
The latest interoperability initiative, HL7’s Fast Healthcare Interoperability Resources, or FHIR (pronounced “fire”), is snowballing this winter. FHIR, based on common web architectures that data geeks refer to as “RESTful,” debuted at HIMSS 2014. But FHIR kicked into overdrive this month thanks to the “Argonaut Project” announced at an HL7 policy conference in D.C. by the JASON Task Force, a joint project of the ONC’s Health IT Policy and Standards Committees. They are working from principles set forth in 2013 and 2014 papers issued by JASON, a group of scientists the Agency for Healthcare Research and Quality contracted to take a fresh look at health data interoperability issues.
What does all that mean in practical terms for healthcare providers, health IT vendors, and ultimately patients? This: JASON advised ONC to mandate health data application program interfaces, or APIs, in stage 3 of meaningful use to address health data interoperability problems. The Argonaut Project hopes to enable that process in a public-private way by providing FHIR implementation guides and specs to the industry by next spring. So far it has enlisted software vendors such as athenahealth, Inc., Cerner Corp., Epic Systems Corp., Meditech, Inc., McKesson Corp., and healthcare providers such as Partners HealthCare, Intermountain Healthcare, and the Mayo Clinic to work on FHIR standards for making health data friendlier to browsers, mobile devices, and (gasp!) competing vendors’ EHR systems. More will likely join; HL7, of course, keeps a running tally of current participants.
Stop me if you’ve heard this story before: Health data interoperability sizzle, which fizzles out in a year or three. The caveat to the FHIR story is that each participant is working on its own use case for FHIR services. There looks to be a lot of work between now and when health data exchange between siloed systems becomes seamless.
But there are flashes of hope. For example, the cross-vendor CommonWell Health Alliance has been working with FHIR to support patient linking and matching services, an excruciatingly difficult data problem to solve in the absence of a national patient identifier (currently, it’s illegal for the federal government to even look into its feasibility). Patients with common names such as John Smith will tell you about the nightmare of tracking their information from one physician to the next, and how it’s difficult to expunge data mistakenly associated with their medical records from other John Smiths. Frustration can lead to misery, when mistaken identity leads to inadvertent termination of health insurance policies or worse yet, running afoul of an employer.
CommonWell’s association with FHIR goes deeper than that. To get McKesson and Cerner to first discuss interoperability a couple years ago, it took a brave soul at one camp to pick up the phone and call his competitor – and then get their respective CEOs on board with a collaboration despite daunting intellectual property hurdles. Those two trailblazers, David McCallie, M.D. and Arien Malec, sit on the JASON Task Force.
In theory, the open APIs the original JASON authors imagined driving interoperable health data will start with the assumption that patients own their data. Not their providers, their providers’ software, cloud vendors, payers, or other third parties. An important second assumption is that the APIs not only will enable patient matching and data security, but also manage permissions set by the patient.
So yes, where there’s smoke, there’s FHIR. John Halamka, M.D., national health data influencer and Beth Israel Deaconess Medical Center CIO, calls it a potential “perfect storm” of interoperability initiative. Even Healtheway, another cross-vendor health data interoperability initiative, “acknowledges growing interest in FHIR” in comments to the HIT Standards Committee despite taking a more skeptical line toward its prospects for success.
What’s your take on the viability and future of FHIR? Is this “the one?” We invite your comments.