Montag, 6. Juli 2020

The "limitations" of FHIR

We have been made aware of the following blog post, which raises the question of whether FHIR is fit for the 21st Century of HealthCare:

A few things...

The article proposes that "...there is a profound limitation of the FHIR standard to be fit for purpose for this interoperability of the 21st Century. That limitation lies in the guideline in development of the FHIR standard specification that calls for data represented in FHIR to be that data which is represented in eighty percent of existing electronic systems."

This is however not the definition of the 80%-rule in FHIR.
The specification says instead: "FHIR resources are designed with the 80/20 rule in mind – focus on the 20% of requirements that satisfy 80% of the interoperability needs. To this end, resources are designed to meet the general or common data requirements of many use cases to avoid the proliferation of numerous, overlapping and redundant resources." 
There is no limitation to preexisting use cases or implementations.

As examples for the limitations of the FHIR Standard, the author lists the the fact that "data such as genomic sequencing, data from the Internet of Things, data from social media and social systems, geolocation data, data from specialized diagnostic systems and from clinical trials and research[...]is all treated as an exception"

Which is an interesting point of view, considering that FHIR has been conceived as a new standard to primarily support integration with web based platforms such as social media and platform restricted mobile devices like wearables and sensor and has grown beyond this initial scope ever since.

Furthermore, genomic data has strong representation in the FHIR world, both in specification and community. 

Clinical studies and research not only features it's own set of core FHIR resources (ResearchStudy and ResearchSubject) but also provides the technological basis for science initiatives such as C3-Pro, Sync4Science or the German Medical Informatics Initiative.

It feels almost redundant to mention the geolocation extension at this point...

The article continues with "perhaps more importantly, the roles of family members and community members who are essential to the care and outcome of an individual as a patient cannot be represented at all. Family members and community members who manage technology that is more advanced than that which existed even in hospitals only 2 decades ago cannot be represented on a patient’s care team."
While scope and usage of RelatedPerson may seem somewhat limited, a quick scan of the list of resources that can reference a RelatedPerson demonstrates, that they can basically perform in any function pertaining to a patient's care, including but not limited to being a participant of the patient's CareTeam.

"It is time to also align our healthcare data standards with the reality of health and care"
I think we need to align this blog post with the FHIR specification, first of all...

...did I mention that we offer FHIR training courses...?