Learning platform in health research and social services

Advancing Research into Learning Health Systems

Clinicians and researchers have great difficulty in quickly and efficiently accessing quality health data, even when they hold the required permission.

This includes data from clinics, hospitals, biobanks, clinical studies, public health, social services, provincial institutes and organizations, wearable technologies (e.g., smart watches), and so on.

It is not feasible to centralize all health data—for social, ethical and legal reasons. Therefore a solution is required that leaves the data where it is while still allowing information for health care and research to be retrieved.

Veuillez accepter le témoin Préférences pour voir ce vidéo.

Dr. Jean-François Ethier, Scientific co-director of the GRIIS, explains how learning health systems respond to the current challenges facing health care and research.

Accessing health data where it is

The PARS3 learning platform is designed to support clinicians and researchers who conduct projects that require authorizations to access data stored in multiple, distributed systems. PARS3 is an integral technological component of a learning health system.

No matter where the data is and no matter how it is managed locally, PARS3 makes it possible to link the clinicians’ and researchers’ queries with the information systems which are specific to each data custodian.

“PARS3 allows data that is distributed across separate computer systems to be processed and networked in a consistent and understandable manner.”

—Jean-François Ethier, Scientific Co-Director of the GRIIS


Related scientific papers

Khnaisser C, Looten V, Ethier J-F, Lavoie L, Burgun A (2023) Combining ontology and temporal databases for data reuse: the example of hospital organizational structures. International Journal of Medical Informatics *** accepted with minor changes Cite
Khnaisser C, Lavoie L, Fraikin B, Barton A, Dussault S, Burgun A, Ethier J-F (2022) Using an ontology to derive a sharable and interoperable relational data model for heterogeneous healthcare data and various applications. Methods Inf Med 61:e73–e88. https://doi.org/10.1055/a-1877-9498 Cite Download
Ecarot T, Fraikin B, Lavoie L, McGilchrist M, Ethier J-F (2021) A Sensitive Data Access Model in Support of Learning Health Systems. Computers 10:25. https://doi.org/10.3390/computers10030025 Cite Download
Dahl LT, Katz A, McGrail K, Diverty B, Ethier J-F, Gavin F, Mcdonald JT, Paprica PA, Schull M, Walker JD, Wu J (2020) The SPOR-Canadian Data Platform: a national initiative to facilitate data rich multi-jurisdictional research. International journal of population data science 5:1–9. https://doi.org/10.23889/ijpds.v5i1.1374 Cite Download
Ecarot T, Fraikin B, Ouellet F, Lavoie L, McGilchrist MM, Ethier J-F (2020) Sensitive Data Exchange Protocol Suite for Healthcare. In: IEEE Symposium on Computers and Communications. Rennes, France Cite Download

Access and exchange procedures used worldwide

The GRIIS collaborates with the research community and with partner organizations at the international level to ensure PARS3 procedures meet the expectations of the world-wide community.

Fast and efficient

Procedures satisfy research requirements.


According to the Operational Data Model (ODM) of the Operational Data Model (ODM) du Clinical Data Interchange Standards Consortium (CDISC).


According to the standards of research ethics boards.


From the point of view of relevant legislation and certification bodies.


From the point of view of our partners, patients and other publics.


An architecture based on Security by Design.


PARS3 is an ecosystem that powers scientific research, knowledge transfer and delivery of health care in learning health systems.

Accurate data queries and results

The PARS3 platform is built on information ontologies (which are similar to lexicons). To make a data request, users use concepts from the ontology that are coherent and understandable from a clinical point of view.

PARS3 then extracts the precisely specified data from the distributed data sources, mediated by other components of the platform. The platform matches extracted data and original query fields using these validated ontological models.

The original source data stays where it is stored — the platform does not generate nor manage data warehouses.

PARS3 responds to health data queries by providing only the minimum data requested as part of a project.

The data is encrypted locally, where it is held temporarily.

The platform itself does not store data and does not interpret content.

The extracted data is subsequently transmitted directly and securely to another secure third party where the requesting user will have access to the data for analysis.


Key partners

With the support of

PARS3 directors

Luc Lavoie - GRIIS - Groupe de recherche interdisciplinaire en informatique de la santé

Luc Lavoie

Co-founder and scientific co-director of the GRIIS

Associate professor of computer science at Université de Sherbrooke

View profile »
Jean-François Ethier - GRIIS - Groupe de recherche interdisciplinaire en informatique de la santé

Jean-François Ethier

Co-founder and scientific co-director of the GRIIS
Full Professor of medicine at Université de Sherbrooke
Research Chair in Health Informatics, Université de Sherbrooke
Licensed physician at the CHUS

View profile »

Improving quality of care and patients’ lives by working together