FHIR ignites new possibilities for interoperability
In 1950, there were 98 TV stations in the United States. Contrast this with 2015 when there were 1,780. (Statista, 2015) Keeping up with relevant information over the years has become an even bigger challenge, with news delivered through email, smart phones, web sites, social media, and other evolving mediums. The good news is that we have technologies like Google Alerts, Digg, Reddit, RSS and the recent addition of Apple News to tightly tailor our consumption to the content we want.
In healthcare, the demands for interoperability – which is the ability of different information technology systems and software applications to communicate, exchange data, and use the information that has been exchanged – grow and evolve with each passing day. For true interoperability to work, we must work together to find ways to connect patient information across the entire care continuum. This means integrating patient records across care systems. It’s a huge challenge, with many U.S. hospitals operating more than 500 different software systems and exchanging around 5 million HL7 messages every day. One hospital saw clinical data traffic volume (measured as the HL7 message count) increase 10 times when it began connecting all the equipment in patients’ hospital rooms.
Like TV in the 1950’s, tuning into the HL7 message stream used to be enough to stay up to date, but the world has changed and today there are simply too many things happening for this “just tune in” approach to work efficiently.
Payment models in healthcare are changing and this is the perfect opportunity to evolve. The focus is shifting from fee-for-service to value-based payment models. With this change, communication of health data across the continuum of care becomes even more critical. The creation of value-based reimbursement contracts and Accountable Care Organizations (ACOs) shift more risk and responsibility for the care of patients onto providers, no matter where those patients may be seeking care. To do that effectively, providers need to know as much pertinent detail about their patients as possible so that information can be used in a meaningful way. This means they need to maximize interoperability.
In 2015, there were 400 hours of video content uploaded to YouTube every minute. (Statista) That means 24,000 people must watch YouTube 24x7x365 just to keep up with what’s being uploaded – and you thought 1,780 channels on TV were a challenge! Fortunately, YouTube created channels and even better, search functionality to help us filter and find precisely what we’re looking for. This isn’t that far off from the shift from traditional HL7 messages to FHIR, which helps providers enter queries to find the relevant information they are looking for.
Clearing up unclear standards
A 2015 KLAS report, titled Interoperability 2015 – Current state and next steps, looked into the state of interoperability. A fair number of provider organizations were interviewed for the report, which states that, “Unclear standards are the greatest hurdle to sharing. Almost no respondents report a lack of standards but instead report frustration with the depth of current standards, as well as poor clarity regarding the application of current standards.” (KLAS Research, 2015)
While there are different types of interoperability out there, for the purposes of care coordination the lion’s share comes in the form of documents. The most commonly referred to specification for the communication of documents is the HL7 Consolidated Clinical Document Architecture (or C-CDA). The good thing about documents is that they can provide an immutable record of an encounter or course of care. For the most part, they represent a snapshot in time of data relevant for the purpose in which the document is being created (e.g. a progress note taken during an encounter). The exchange of these documents can give providers insight in to their patients’ encounters and episodes of care. There is value in the exchange of these documents, particularly in use cases where a patient is being transitioned into another care organization.
Documents can contain a lot of information about a patient. In fact, they may contain too much information. In a recent meeting of the Interoperability Experience Task Force of the U.S. Federal Advisory Committees for Healthcare IT Policy and Standards, Scott Stuewe of Cerner noted that, “Future exchange will be challenging because there will be too much data in too many documents from too many structures without standardization.” (IXTF, 2016) What is being stated here is that there is just too much information that providers do not need or is being delivered in a way that they can’t effectively use in a typical 15-minute patient encounter. As we move into an ever more connected world with new sources of data from emerging areas like genomics and Patient Generated Health Data (PHGD) from health apps and FDA regulated medical devices alike, this data volume concern will grow exponentially.
The rise of APIs and consumer empowerment
Information access has changed. Gone are the days of getting pushed all your news and information from that handful of TV stations and a single newspaper subscription. We now live in a consumer driven data market because we can seek out the information we want. As we mentioned earlier, the amount of information and the number of data generators continues to rise exponentially, and with that, there are tools that help us manage that complexity, filter it, and make sense of that data so that we can get the information we are looking for. Whether you may realize it or not, much of this data is handled by numerous systems using web-based Application Program Interfaces or APIs.
The increase of data is no different in healthcare. The amount of data creation is growing and will only continue to accelerate with advances in things like research, genomics, and health monitors. The predominately used standards in healthcare today simply weren’t designed to address a lot of the emerging needs.
The good news is that the HL7 Fast Healthcare Interoperability Resources (FHIR – pronounced “fire”) standard brings an API approach to healthcare interoperability. FHIR is based on the same protocols and interaction patterns that the Internet is built on and does so in a way that enables many kinds of healthcare related apps to more readily connect to one another. It enables a data consumer driven approach – similar to YouTube – that empowers consumers to actively submit, query, and retrieve the data that they want in real time, rather than passively accepting what was handed to them. So, if you are only interested in data specific to your needs that can be integrated into your workflow in real-time, a FHIR enabled world is just what you are looking for.
In a secure fashion, FHIR supports an unbounded array of use cases from traditional clinical care to analytics and population health management to direct consumer engagement and empowerment. It enables system developers to create a platform in which data can flow much more easily, and securely, beyond the traditional healthcare settings and systems. And, unlike C-CDA, it is, from its very point of conception, written in a way that is meant to be understandable to the implementer; that is to say, understandable to the systems developers that will create the rich healthcare application ecosystem. Easily understood and accessible standards represent a lower barrier to adoption and are critical to ensuring proper and consistent data expression that all systems will be able to process.
Weaving FHIR into the interoperability tapestry
FHIR was designed in a way that can address the issues identified in the KLAS Interoperability report regarding unclear specifications, as well as many of the primary challenges that the Interoperability Experience Task Force have identified.
With all the promise however, a word of caution must be provided. FHIR, while it is gaining traction, is still maturing. As tempting as the idea of an interoperability silver bullet is, no such thing exists. We need to remember that FHIR is just one thread in the complex tapestry of healthcare data interoperability. Ultimately a successful interoperability strategy will require the ability to weave together a wide-ranging set of data sources and formats into a cohesive solution.
Just like television and YouTube have evolved and grown exponentially in their content production – and will continue to do so – to meet consumer demands, healthcare standards, including FHIR, will need to adopt the same flexibility to meet the ever-increasing amounts of big data and the evolving requirements of interoperability.
* * *
The real world on FHIR
Hackensack UMC (Hackensack UMC) is the fourth largest hospital in the United Sstates by admissions. Recently, Hackensack UMC took the bold step of asking their patients for ways the medical center could make their care experience better. Not surprisingly, patients reported frustration with wait times, repeatedly filling out the same paperwork, and data inaccuracies. Hackensack UMC also met with nurses and providers to understand their ideas and suggestions.
By leveraging their interoperability platform, Infor Cloverleaf, to expose a FHIR interface for their backend clinical systems, Hackensack UMC was able to vault past the hurdles of disparate protocols, paper, people, and systems to reinvent the patient experience. With this contemporary API platform, they were able to build mobile applications that deliver a smooth, paper free experience to the patient – all enabled by FHIR. This is possible even though behind the scenes there exists a mix of systems, data, and transmission protocols.
A new patient who wants to schedule an appointment can use a mobile device to enter demographic information, choose a physician and schedule appointments. Once appointments are confirmed, the patient receives a secure text message asking him or her to complete a questionnaire, which can be completed via mobile device. The patient checks in for the appointment – without the need to fill out additional questionnaires or provide additional information. Cloverleaf pulls information from several electronic medical records and other data repositories, which provides Hackensack UMC with patient snapshots for doctors and nurses before the patient arrives. That gives the patient and provider more time to focus on a discussion about care – and helps improve patient care and satisfaction.
C-CDA, Consolidated-Clinical Document Architecture, Fast Healthcare Interoperability Resources, FHIR, HL7, Infor, interoperability, Interoperability Experience Task Force, KLAS, Patient Generated Health Data, PHGD