2018 is the year that FHIR has established itself as the de-facto current and future standard for healthcare data exchange, seamless integration, and interoperable applications. 2019 promises to be even more exciting for developers and providers.  Here are some major milestones:

  • Apple releases its HealthRecords app for iPhone and iPad and embeds FHIR into iOS
  • Provider-centered DaVinci project launches with participation from industry heavy-hitters like Optum, Cerner, Epic, Anthem, BlueCross and Humana
  • Tech giants endorse FHIR including Amazon, Microsoft, Oracle, Google and SalesForce
  • US Federal government official Healthcare IT agency ONC demonstrates increased support for FHIR, releases Inferno Testing Suite
  • US government agency CMS (Medicare, Medicaid) releases FHIR-based BlueButton 2.0 initiative providing access to data for 53M beneficiaries
  • Major US EHR vendors including Epic, Cerner, Allscripts, expand developer access to FHIR-based API’s

Blockchain is a distributed ledger technology with built-in strong encryption that enables permissioned sharing of data within a defined ecosystem.  Blockchain, in tandem with artificial intelligence, has the potential to radically transform the healthcare industry by ensuring data fluidity within the ecosystem and allowing meaningful insights to be extracted from overwhelming volumes of big data.Numerous use cases are currently being explored for blockchain technology within healthcare, affecting payers, physicians and patients in fundamental ways.  For payers, blockchain promises increased trust in data sharing with reduction of errors and enhanced efficiency in administrative processes.  For physicians, AI can be used to discover actionable insights in data, for example in automating analysis of medical images to screen for abnormalities, allowing providers to focus more on critical tasks and decision making.

Patient engagement with their own healthcare may also be enhanced with blockchain, for example, if patients are able to verify that their records are complete, current and accurate on a blockchain, AI algorithms can be applied to stratify risk, suggest diagnoses or identify gaps in care.

Big Data in healthcare has been defined in terms of ‘3 V’s of volume, velocity and variety. The reality of the health industry today is that the volume of data is now measured in exabytes (billions of gigabytes) and this data is increasing (by 2020) 50% per year.  Big Data derives from the digital transformation of healthcare which has ushered in the era of electronic health record systems (EHRs) along with technological advances such as digital medical imaging, precision medicine and genomics, all of which serve as drivers to an ongoing exponential data expansion.

This explosion of digital data has outrun the capacity of the existing tools in the healthcare armamentarium to exercise effective data management.  As a result, new tools and technologies, developed for other industries to manage big data, are rapidly being imported into healthcare.  One example is data storage where traditional relational databases are being replaced by a new generation of ‘no-sql’ databases such as Hadoop and MongoDB.  Another illustration involves artificial intelligence which has a synergistic relationship with big data:  effective AI algorithms generally need to be trained on very large datasets and, once trained, can be applied to extract meaningful insights from the data.

In a recent survey more than ¾ of healthcare organizations agreed that Internet-of-Things will be transformative to the industry, bringing about increased innovation and a massive reduction of costs.

The IoT has numerous applications in healthcare, from remote monitoring to smart sensors and medical device integration. It has the potential to not only keep patients safe and healthy, but to improve how physicians deliver care as well. Healthcare IoT can also boost patient engagement and satisfaction by allowing patients to spend more time interacting with their doctors.

But healthcare IoT has several major challenges, starting with the volume of big data associated with connected devices.   Management of this data and integration with existing workflows is imperative as is implementing additional security.

Legacy applications, e.g. HL7 v2, CDA .. may have to be converted to FHIR for purposes of integration or interoperability.  This process typically involves the use of software transforms which perform the translation from legacy data to the new application requirements.

Integration engines provided by numerous vendors are suitable for these tasks although the implementations can be time-consuming and costly.  (Some vendors offer graphical drag-and-drop user interfaces that can be more efficient for building the transform templates.)

More recently, HL7 has released a tool called FHIR Mapping Language (FML) that promises to expedite legacy to FHIR integrations.

FHIR-based applications are agnostic with respect to what backend database technology is utilized.  Some FHIR platform vendors have chosen to work with traditional RDBMS tools, e.g. Health Samurai which has built its proprietary FHIRBase db on top of Postgres.  Others, including Helios and WBM Health, have elected to implement Big Data no-sql data stores (Apache Cassandra and MongoDB, respectively).

FHIR will be a transformative technological advance throughout the entire healthcare industry.  Fields affected include:

  • EMR and EHR Vendors
  • Healthcare Data Interchange Solution Providers
  • Claims Systems
  • Pharmacy Systems
  • Medical Terminology Systems
  • Other existing solution vendors

HL7 FHIR provides an open standard API to enable interoperable sharing of medical data and to encourage innovators to create apps that seamlessly and securely run across the healthcare system.   Commercial vendors as well as nonprofit initiatives have endorsed the FHIR standard and built diverse platforms on top of it that facilitate open and highly integrated applications.

One such initiative that has gained widespread popularity is SMART-on-FHIR which originated in US universities.  The SMART project has been endorsed by the major EHR vendors such as Epic and Cerner, thereby ensuring that apps created within the SMART framework will integrate seamlessly with the EHR.

The kinds of apps that can be built using FHIR and that integrate with the EHR span a wide spectrum as can easily be seen by browsing the App Gallery on the SMART website;  these cover categories  including Clinical Research, Data Visualization, Genomics, Medication, Patient Engagement, Disease Management, and Population Health.

Until recently, most of those working in healthcare IT would subscribe to the former assertion, but some influential voices have suggested an alternative view.

Steve Munini, founder of Helios Software, has described a number of different FHIR-based architectural approaches including:  Interoperability Interface, FHIR Broker Adaptor, FHIR API Encapsulating a Vendor-Neutral Clinical Reposity and more.

Munini concludes that there are two fundamentally different architectural premises to consider:

The first is adapting a solution to fit FHIR by providing FHIR as a new interface on top of an existing solution (the most common approach).  The second approach is to make use of the richness and expressiveness implicit in the FHIR specification to create FHIR-native solutions.