During the Middle Ages, London’s Bartholomew Fair grew from what we would call a trade show for cloth merchants into a full-blown annual exposition where merchants could conduct business in the presence of “sideshows, prize-fighters, musicians, wire-walkers, acrobats, puppets, freaks and wild animals.” HIMSS’15 in Chicago next week promises all that and more — hopefully without the wild animals. This year at HIMSS, I hope to talk to a lot of different companies and people about interoperability.
Since the publication of our CNM Market Trends Report at the end of last year, it has become clear that the market is seeking interoperability capabilities from EHRs and HIE/CNM that look remarkably similar. As the system of record within an HCO, the EHR is potentially current. As the system of record for the community-wide longitudinal record, the HIE/CNM is potentially comprehensive. These resources could, if they were rendered more accessible, provide the patient data to fuel a vast array of point-of-care, population level, care management, and administrative applications to help the healthcare system shed its fee-for-service blinkers.
Interoperability-related developments of the last few months have set the stage for my HIMSS’15 discussions: the MU3 NPRM, the trial balloon for ONC’s Interoperability Roadmap, a dramatic uptick in interest in HL7 FHIR, the establishment of the Argonaut Project. Healthcare’s interoperability issues have leaked out into general press coverage. The line between genuine interoperability issues and the heat generated by the machinations between EHR vendors and contractors over the impending DHMSM acquisition is cloudy. But radically improved interoperability is a pre-condition to a healthcare system that is patient-centric, population focused and value-driven.
A technology featured in all of these developments is API-based access to application data. This approach is old hat in every part of the economy except healthcare. Walgreens has an API to its photo print operations as well as for its clinic appointments and prescription refills. Bloomberg has APIs to its market data. Twitter provides an API to its vast store of tweets. Virtually any customer-facing operation has APIs for third party developers. Fueled by the rapid adoption of smartphones, enterprises are using RESTful APIs and Open Authorization (OAuth2) to capitalize on the value of their data to drive revenue, lower costs, and better serve customers – B2B or B2C.
The key word here is capitalize. Businesses are charging other businesses for access to this data. In a perfect world, every HCO could provide an API to its EHRs and other applications to support care coordination and population health and … you get the idea. Obviously, this was the idea behind HIE and its predecessor concepts. Proposed legislation circulating in Congress contains language that would require APIs as a condition of certification for EHRs. APIs have captured attention in the industry. They have so far not captured much market share.
APIs are not really all that new in healthcare. CarePass was a cloud- and API-based effort by Aetna to make data from applications and wearables more widely available to patients via applications from third-party developers. It closed at the end of 2014, a victim of patient and clinician disinterest and mixed support within Aetna. Despite this example, I think that broad-based API-based access is inevitable in healthcare. Lightweight, discrete data access can make exchange more financially palatable than the all-or-nothing datacenter-scale HIEs that struggle with funding. This year at HIMSS, I will be paying close attention to what vendors and HCOs are saying about APIs and their potential uses.
HL7’s FHIR will also be big in Chicago this year. It will be a set of Representational State Transfer (REST) web services APIs. It promises an easier-to-program, less resource-intensive alternative to the more widely deployed SOAP-based IHE profiles. It also promises discrete data access to supplement the various methods for document-centric data access now widely deployed around the healthcare system. I fully expect to see a lot of flame imagery around the McCormack Center this year as well as newly created FHIR marketing collateral. I hope also to find out about some concrete plans for incorporating FHIR profiles into existing EHR and HIE product functionality.
I also want to know more about Project Argonaut, which aims to move the industry toward widespread use of FHIR and APIs. This standards effort proposes to make rapid progress on use cases for clinician-to-clinician exchange and consumer access. How one define “rapid progress” in this industry is difficult to judge. For example, CommonWell Health Alliance launched at HIMSS’13 and only recently published its specification.
Eventually the powers-that-be shut down the Bartholomew Fair because it encouraged debauchery and disorder. HIMSS shows no such inclinations but I plan to talk to as many EHR and HIE/CNM vendors as possib
le and hear both legitimate and outlandish claims about APIs, FHIR, and better ways to support a more interoperable HIT infrastructure.
Interoperability is a pipedream. There’s already plenty of “interoperability” because every installation of a clinical system requires integration with _something_ else. Even if that’s a practice management system or an accounting system, data is moving in and out of a CIS to some degree. Is it easy? Is it cheap? Is it ‘standards based’? Hell no. But that’s what keeps everyone in business. Is the huge professional service fees needed to make “off the shelf software” work in any implementation.
The fastest path to a ubiquitous comprehensive clinical record that is accessible to all providers for all patients is Blue Button. You’ll never see all provider entities connected to all other provider entities for ubiquitous access to records. But the patient is present in all encounters and has a larger vested interest in managing the information from point of care to point of care than the provider does in integrating with her competition.
Blue Button? The ubiquitous answer to everything. Peter Levin would be proud of you, Scott. It will even make my tea and crumpets in the morning. Earl Grey, if you please.
We will agree on one point, good sir. That being that 100% interoperability is a pipe dream, as you say. It requires, at minimum, the ability for all concern to define what that is at a policy level then create some detailed model accepted industry-wide on what that true means.
Blue Button has it’s place and that’s getting information to a select group of patients and caregivers. For that, I see great value as just one of many standards that you can use. As a geek clinician very familiar with the promise and limitation to Blue Button, it won’t pass muster for me to use. It’s all about the workflow and the richness of the data, neither of which Blue Button can help me with.
Cheers! Ever forward.