JMIR Publications

ABSTRACT

Background:

The Fast Healthcare Interoperability Resources (FHIR) standard is now well established as a global standard for healthcare information exchange. The value of FHIR is also being actively explored to support clinical and healthcare research in areas such as observational studies using “real world data” (RWD) and in clinical studies intended for regulatory submissions where it plays a key role in enabling direct data capture from clinical sites. The goal of the work for clinical trials is to use FHIR resources to define study schedules of activities (SoA) to support operational implementation at research sites. A HL7 FHIR Implementation Guide (IG) has been published that can define basic SoAs, but this does not offer solutions to define some types of study scheduling (e.g. treatment cycles) or other scheduling requirements (e.g. conditional switching such as that found in vaccine studies)

Objective:

The objective of this work was to re-investigate how the FHIR definitional resources, particularly PlanDefinition might be extended or revised to enable the model to cover the wider range of SoA scheduling requirements that are commonly encountered in clinical research studies.

Methods:

A previously described SoA graph model was used to investigate, extend and test the fundamental requirements for SoA definitions modelling the complex and conditional requirements outlined above. These requirements were then defined/reflected as FHIR definitional resources with, where necessary, FHIR permitted technical solutions (extensions, value sets, etc.). Specification accuracy was tested by comparing the SoA graph model attributes and relationships with those able to be recovered from the FHIR specifications.

Results:

Using confirmed graph-based models of clinical SoAs, a SoA FHIR model based on the PlanDefinition resource was developed that can model simple, complex and conditional SoA requirements for a wide variety of SoA use cases. Using example and publically available study SoAs, the SoA FHIR model was able to accurately specify a large range of commonly met study scheduling requirements.

Conclusions:

A FHIR model for the specification of clinical trial SoAs has been developed that offers a more comprehensive set of scheduling requirements to be defined than previously considered. Particularly it implements methods for specifying conditional scheduling requirements, and clearly separates SoA activity definition from the study scheduling requirments.