February 19, 2019 | Julian
Access to the right data, at the right time, for the right person has always been a big challenge in healthcare. Whether that be a clinician accessing a patient record at appointment time, or a researcher requiring access to a large anonymised dataset for analysis as part of a project.
There are many complexities which make it difficult to provide access to the required data. Ultimately this complexity leads to data unavailability, delays and an overall inefficient process.
Fundamental problems include:
Different data models everywhere!
Over time different hospitals and departments have stored their data in very different ways. This disparity, or lack of standardisation, means that complex data integration efforts are required to bring in new systems, link data between existing systems and often even to fulfil data requests.
Existing standards, such as CDA, are unwieldy and require significant upfront investment to define a plethora of data points. Furthermore adoption of standards is not uniform and healthcare providers are faced with many legacy databases. Data integration requires data specialists, is time consuming and thus expensive; this is a barrier to unlocking insights into the data which is key to pushing the industry forward.
FAIR data access
Whilst healthcare understandably has strict processes in place to control access to data, this often means that “outsiders” have no way to easily discover what data is available – stifling research. To promote innovation, data needs to meet the FAIR standards, that is findability, accessibility, interoperability, and reusability.
Help is at hand from FHIR (Fast Healthcare Interoperability Resources)!
What is FHIR?
FHIR is emerging as the preferred standard for exchanging healthcare information. HL7 (Healthcare Level 7) is the not for profit organisation behind the FHIR standard; their vision is for “A world in which everyone can securely access and use the right health data when and where they need it.”
Resources are the key building blocks of FHIR
Resources make FHIR modular, enabling implementers to focus on only the areas of relevance, thus avoiding some of the pitfalls of existing standards.
“Think of Resources as paper “forms” reflecting different types of clinical and administrative information that can be captured and shared. The FHIR specification defines a generic “form template” for each type of clinical information – so one for allergies, one for prescriptions, one for referrals, etc.” – hl7.org/fhir/overview-clinical.html
The FHIR community, via numerous working groups, is working together to define Resources that are applicable globally. A key principle behind Resource definition is the 80/20 rule, which focuses on the 20% of requirements that satisfy 80% of the interoperability needs. FHIR accepts that there will never be a one-size-fits-all or works out of the box solution and focuses on minimising the effort required to integrate systems by agreeing on common data elements, their groupings and by providing a framework for extensibility and customisation, i.e. the 20%.
FHIR’s rapid increase in adoption was spurred by the US Office of the National Coordinator for Health Information Technology (ONC) issuing “certification criteria” which included guidance regarding use of APIs for health data exchange. This led the community to select FHIR, which uses modern standards for data access, data manipulation (RESTful APIs), authentication (OAuth), and data representation (XML, JSON, HTTP etc). The use of these modern standards is key to enabling efficient development of health applications (e.g. mobile, and web apps), furthermore the door opens to wider proportion of software developers who are already familiar with using these technologies.
FHIR DevDays conference
In November I was lucky enough to attend the excellent FHIR DevDays conference in Amsterdam where I was able to hear first-hand the ways in which FHIR is helping to address the problems outlined above. The FHIR community is a diverse mix of healthcare professionals, technical specialists, pharma, EHR system providers (e.g. EPIC, Cerner) and cloud service providers / big business (e.g. Microsoft, Google). There was a focus on patient empowerment, and enabling patients to have access to and be in control of their own data (check out the e-Patient Dave ted talk) – an area in which FHIR is helping paving the way. Overall there was a buzz in the air, and it feels like FHIR is nearing a tipping point of adoption (particularly in the US) resulting in positive change.
FHIR is here to help
As FHIR resources mature and more organisations and system providers adopt FHIR for data exchange, integration between healthcare systems will become simpler. Imagine 80% of your integration work was done for you straight out of the blocks, that’s one goal of FHIR!
FHIR adoption is increasing. A recent article I came across stated that greater than 50% of new healthcare projects in the US are now using FHIR. Although FHIR is more prevalent in the US, it’s spreading – FHIR has affiliates in 35 countries. On the 1st of January 2019 FHIR Release 4 (R4) was published which is the culmination of 18 months effort, with R4 comes the first “normative” content, future iterations of normative content being backwards compatible. It’s plain to see how FHIR is addressing the issue of disparate data models.
With regards to FAIR data access, we have seen the effect that governmental pressures can have (e.g. ONC’s certification criteria) in forcing data owners to re-think their sharing model. This in turn has driven adoption of new FHIR enabled systems, which increases data visibility by publishing available resources. A public data catalogue integrated with internet search engines such as Google Dataset Search can let potential researchers know what data is being published in an API like a FHIR service.
FHIR at Aridhia
Aridhia provides services for hospitals to interface with our collaborative research platform, sharing data with a view to driving research and testing applications integrated with clinical workflows. FHIR provides a useful standard for relaying Electronic Patient Record (EPR) data for analysis and SMART on FHIR provides a convenient standard for us to facilitate our customers to integrate with third-party EPR systems. To date we have been building up our capability in this area and look to invest more in new services as customers adopt the standards.
Last year we partnered with Microsoft, and in November Microsoft released their FHIR Server to the community, which we look forward to exploring further as we transition our services to Azure. Microsoft FHIR Server is just one of the wider sets of Azure services that we’ll be able to offer as we continue to provide a world leading digital research environment that enables our customer to perform healthcare data research at scale.