LogoLogo

OSIRIS

Overview

OSIRIS (Interoperability and data sharing of clinical and biological data in Oncology) is a French national minimum data set and Common Data Model for clinical and genomic data sharing in oncology, developed through a bottom-up national consensus process launched in 2015. Funded by INCa (the French National Cancer Institute), it was built by a task force of Integrated Cancer Research Sites (SIRICs), comprehensive cancer centres (Unicancer network), and university hospitals. The resulting model specifies 67 clinical items and 65 omics items in an event-based data model that captures longitudinal disease course, treatment, and genomic findings in precision medicine trials. OSIRIS is aligned with HL7 FHIR (a full FHIR R4 Implementation Guide exists), approved by a National Plan Strategic Committee, and freely released.

Scope and Model

The OSIRIS event-based data model organises data around Treatment Plan Episodes (TPEs), clinical moments anchored to treatment decisions. It captures patient characteristics (demographics, family history, clinical trial participation), disease characteristics (ICD-O-3 tumour histology, ICD-10 diagnosis coding), treatment (drug administration, surgical procedures, radiotherapy, adverse events), genomics (sequencing analysis, SNVs, copy number variants, fusions, annotations), response evaluation (RECIST, biomarker measurements), and imaging (radiomics, imaging study metadata).

Key Features

  • The model is modular and extensible: the core oncology model can be extended to specific cancer types.
  • It supports retrospective and prospective data from EHRs, eCRFs, data warehouses, and cancer registries.
  • A FHIR Implementation Guide (v1.1.0) is published at ig-osiris.cancer.fr.
  • The model is GDPR-compliant by design and has been reviewed by CNIL.
  • A proof-of-concept was demonstrated on a 300-patient cohort across multiple French cancer centres.

Connections

Resources

Graph View

Tags

Backlinks