Francesco Torelli
Engineering - Ingegneria Informatica Spa

The main goal of CrowdHEALTH is to provide a set of integrated tools for supporting the development and evaluation of health policies. A key element of the CrowdHEALTH platform is the so called Holistic Health Record (HHR) model. It’s called “holistic” because it is a data model not limited to clinical information, but that represents different health-related aspects of the citizens.

Personal Health Record (PHR) models are focused on wellness relevant information collected by the citizen. Numerous research initiatives proposed PHR models and cover topics such as allergies, immunizations, medication, home monitoring, genetic features. Standards for health data exchange, such as CCR, DICOM, HL7 CDA, HL7 RIM,HL7 FHIR, openEHR, CEN/ISO EN13606 covers more clinical information produced by health professionals and are exploited by health institutions worldwide for their EHR (Electronic Health Record) systems. Specific health data models, have also been defined for representing information used for or produced by medical researchers, such as HCSRN’s Virtual Data Warehouse (VDW) and Observational Medical Outcomes Partnership (OMOP) model.

CrowdHEALTH project follows the FHIR principles, but also developed a new approach for the definition of FHIR profiles and the HHR model has been specified using this new approach.

While FHIR is intended for the exchange of health data, especially clinical data, among different application contexts, CrowdHEALTH focuses on a more restricted kind of application that is: the integration within a same context of holistic health data. For this purpose CrowdHEALTH defined a specific profile corresponding to the HHR model.

CrowdHEALTH defined a FHIR profile that:

  1. adds new kind of information not included in the basic standard, especially non clinical data, such as social data, personal nutrition data, lifestyle and fitness data provided by the citizen.
  2. constraints the FHIR data schema, so that each kind of information may be represented exactly in one way.

The second point is especially important to assure that data coming from different data sources may be coherently integrated. Because of its generality, the FHIR standard gives the freedom to use different kind of resources and different terminologies to represent the same information. In contrast, the profile defined within CrowdHEALTH imposes to represent each different type of information with one specific term from one specific terminology and one specific property of one specific type of resource. This reduces the ambiguities and possibility of errors, and simplifies the data integration tasks, especially when different independently actors populates a same integrated database using different sources.

Future research will be focused on further development of a new approach for the specification of other FHIR profiles.