What is a FHIR data store?
A FHIR data store is a clinical data repository (CDR) that aggregates and stores healthcare data (demographics, results, images, admissions, transfers, diagnosis, etc.) from multiple source systems into a single location in HIPAA-eligible HL7® Fast Healthcare Interoperability Resources (FHIR®). Once established, a FHIR store can serve as the foundation for real-time analytics, reporting, and data sharing.
Why FHIR?
While FHIR was designed for data exchange, not persistence, there are a variety of reasons why storage in FHIR makes sense:
FHIR is highly flexible and interpretable in a wide range of systems and settings:
Most legacy healthcare standards can be translated to FHIR.
FHIR is easily shared across systems.
FHIR is human-readable.
The shape of FHIR data in stored form is the same as it is in transit
FHIR enables modern use cases:
FHIR supports the ability to query, retrieve, and dissect resources - a prerequisite for application in artificial intelligence (AI) and machine learning (ML).
FHIR future-proofs compliance:
Regulatory bodies like CMS and ONC have pushed for FHIR adoption in many recently finalized and proposed rules.
All major hyperscalers, including Amazon Web Services, Google Cloud, and Microsoft Azure have selected the FHIR format as the standard for clinical data ingestion and storage.
“FHIR is the gateway to AI and machine learning.” - Aashima Gupta, Global Head of Healthcare and Life Sciences, Google Cloud
Why build a FHIR store?
A FHIR store allows organizations to quickly turn on new use cases as business opportunities arise, compliance mandates are issued, or technology advances. By translating and normalizing healthcare data from various disparate sources into a single, secure FHIR store, organizations will enable their first intended use case, while also building a foundation that can effectively serve and accelerate all future use cases.
Adopt artificial intelligence and machine learning
The opportunities for cloud-based AI, ML, and other advanced analytics tools to transform healthcare are innumerable, but if we enable them use case by use case, it may take decades to fully expose their transformative power. Building a FHIR store is the best way to accelerate the use of AI/ML in healthcare.
Simplify quality measure submission
NCQA is piloting the FHIR data model for digital HEDIS quality measures. A FHIR store can streamline the process of collecting, normalizing, and submitting quality data. Once in a FHIR store, data can be checked against CQL logic and submitted by a certified NCQA data partner. It is expected the use of FHIR for quality measures will expand in the coming years, and if pending CMS rules are finalized, it may even be required. Building a FHIR store now is the best way to make sure your organization is prepared, with the added benefit of reducing human error, bias, and administrative burden.
Create longitudinal patient records
Because most legacy healthcare standards can be translated and normalized within a FHIR store, organizations can finally achieve their vision of enabling cross-system longitudinal patient records. These records can then be used as a backend by other products and applications to enhance patient engagement, manage population health, improve care, and more.
Enable app/product development
FHIR is a ground-breaking standard because it is programmable. Builders can do more with FHIR than was ever possible with HL7v2 or other legacy standards. A variety of tools have already been developed on the FHIR spec (TestScript resources, Microsoft CodeGen, SQL on FHIR, and many more). A FHIR Store, paired with cloud services and technologies is the gateway to incredible innovation.
Stay ahead of compliance and quality requirements
The ONC and CMS have shown their intention to accelerate FHIR adoption through recent interoperability mandates and proposals. Getting data in one place and in FHIR positions healthcare organizations for compliance as regulations are finalized and enforced.
The challenges of building a FHIR store
While the advantages are clear, building a FHIR store comes with challenges that can be daunting for the majority of healthcare organizations today.
Because most healthcare data is in legacy standards (HL7v2, CDA, X12, DICOM) translation is necessary. Because there are nuances in the implementation of every standard, translating and normalizing data to FHIR can be cumbersome and overwhelming at high volumes. Our recent research found that organizations struggle to translate and enable data ingestion into the cloud due at scale to challenges with:
Human capital
68% of organizations lack the necessary human resources and/or skillsets
Financial capital
52% of organizations lack the budget required
Technical capital
44% of organizations lack the technical infrastructure
As a result, projects are lengthy and costly. Generally they:
58% require a system integrator
86% take longer than 6 months to complete
65% cost more than $500k to implement
How Redox helps
Redox can accelerate the creation of FHIR stores by up to 80% faster than doing it yourself or enlisting help from system integrators. Nova securely extracts data from legacy systems, translates, and normalizes to FHIR for easy onboarding into AWS, Databricks, Google Cloud, or Microsoft Azure FHIR stores. We achieve unmatched speeds by leveraging a library of over 4,000 configurations that we’ve cultivated over 5,000+ healthcare integrations.
Nova currently supports FHIR R4, including support for US Core profiles based on R4.. This includes support for a growing number of FHIR extensions. Nova currently supports the following translations:
C-CDA to FHIR: supports the parsing and rendering in FHIR of over 42 C-CDA section templates, covering 29 resource types (so far).
HL7v2 to FHIR: supports 21 resource types across the most common HL7v2 event types (ADT, SIU, ORU, etc).
X12 to FHIR: planned support for 15 resources across a selection of X12 events.
DICOM to FHIR: supports the retrieval of images and parsing of metadata to FHIR from legacy DICOM standards.
Additional details on our translation capabilities and processes can be found in our Dev Docs. We are constantly expanding and refining Nova translation capabilities as we work with our customers and through active participation in FHIR standards development. We are a proud annual sponsor of HL7 FHIR Dev Days and a regular participant in HL7 Connectathons.
Nova is built for extreme flexibility so no unicorn is out of bounds. If your organization or data sources have unique requirements, we have a full range of professional services available to deliver what you need.