Baseline
Initial data, perimeter, population and full costs.
Legal file · 5.07
5.07 — Early Disease Detection Using Artificial Intelligence
The Plan presents this as a structural regulatory measure. A system influencing diagnosis or care may fall under medical-device rules and the EU AI Act. Clinical benefit must be demonstrated before broad deployment.
What the Plan proposes
Plan source: the measure title, stated effect and evidence level come from the public Plan de Rupture. This page adds an initial legal and operational architecture requiring expert review.
The Plan presents this as a structural regulatory measure. A system influencing diagnosis or care may fall under medical-device rules and the EU AI Act. Clinical benefit must be demonstrated before broad deployment.
Status: structured citizen prototype requiring review by ministries, the Conseil d’État, professionals, patient representatives and competent authorities.
Implementation path
- Define intended purposeScreening, alerting, diagnostic support and prioritisation are different uses.
- Validate clinicallyTest on representative data against an appropriate medical reference.
- Measure errorsPublish sensitivity, specificity, false positives and false negatives.
- Inform and superviseProfessionals and patients should understand the system’s role.
- Monitor after deploymentDetect drift, incidents and unequal performance.
What must be measured before and after
Quality of care
Safety, continuity, delays and patient outcomes.
Workforce effects
Training, workload, mobility, attractiveness and working conditions.
Net effect
Gross savings minus transition, recurring costs and shifted burdens.