LUCC lung – Large and Unified Cancer Cohort

20/10/2026
20/10/2026
Data source
Human
Hospital discharge records
Hospital inpatient records
Hospital outpatient visit records
Primary care medical records
Download as PDF
Access and validation

Governance details

Documents or webpages that describe the overall governance of the data source and processes and procedures for data capture and management, data quality check and validation results (governing data access or utilisation for research purposes).

Charte Comité scientifique LUCC (vfinale).pdf

English (264.25 KB - PDF) View document
https://www.lifen.fr/database/lucc-poumon

Biospecimen access

Are biospecimens available in the data source (e.g., tissue samples)?

No

Access to subject details

Can individual patients/practitioners/practices included in the data source be contacted?

Yes

Description of data collection

Medical records are collected from participating hospitals either automatically through Lifen interoperability connectors (nominal flow) or, for non-connected sites, via Datalab Bridge, a free web application that pseudonymises documents before transfer to a secure SFTP server.
Documents are pseudonymised before ingestion into the clinical data warehouse (EDS) and re-pseudonymised with a project-specific UUID in each project workspace. Variables are then extracted from unstructured text by a fine-tuned open-source Mistral AI model, with targeted human validation by clinical research professionals (hybrid AI-human workflow).
The variable set is defined and validated by the cohort's scientific committee. Hosting is HDS-certified (health data hosting, France).
Event triggering registration

Event triggering registration of a person in the data source

Disease diagnosis

Event triggering de-registration of a person in the data source

Other

Event triggering de-registration of a person in the data source, other

Exercise of the right to object
Data source linkage

Linkage

Is the data source described created by the linkage of other data sources (prelinked data source) and/or can the data source be linked to other data source on an ad-hoc basis?

Yes

Linkage description, pre-linked

None. The data source is not created by the linkage of other data sources: it is built from medical documents collected directly from participating hospitals

Linkage description, possible linkage

Ad-hoc linkage with French administrative healthcare databases (SNDS) is considered, in particular as a strategy to recover missing data such as vital status.
Data management specifications that apply for the data source

Data source refresh

Monthly

Informed consent for use of data for research

Not Required

Possibility of data validation

Can validity of the data in the data source be verified (e.g., access to original medical charts)?

Yes

Data source preservation

Are records preserved in the data source indefinitely?

No

Approval for publication

Is an approval needed for publishing the results of a study using the data source?

Yes

Data source last refresh

Common Data Model (CDM) mapping

CDM mapping

Has the data source been converted (ETL-ed) to a common data model?

Yes