The French National Health Database: An opportunity to produce real-world evidence for the Early Access Program?

A key pillar of the EAP is the requirement to collect real-world data (RWD) through the Protocole d’Utilisation Thérapeutique et de Recueil des Données (PUT-RD). Designed to capture critical clinical and demographic information on patients benefiting from early access, the PUT-RD relies on hospital clinicians to complete the forms, with financial compensation provided by the sponsor.

The PUT-RD currently requires the collection of a wide range of variables, including patient characteristics, medical eligibility and non-eligibility criteria, and disease-related data, such as diagnosis, prior treatments, and comorbidities. Data on treatment, including dosage, concomitant treatments, and temporary or permanent treatment interruptions, are also required. Efficacy data must be adapted to the specific medication and include both objective criteria and, when applicable, patient-reported outcomes (PROs) such as quality of life or symptom assessments using standardized questionnaires. Additionally, safety and tolerance data, including adverse events and special situations, are collected throughout treatment. These data are gathered at multiple points, including treatment initiation, follow-up, and upon definitive treatment cessation.

The data collected through the PUT-RD are crucial for evaluating treatment efficacy, safety, and real-world use. They inform decisions by the HAS and the French National Agency for the Safety of Medicines and Health Products (ANSM) regarding EAP renewals, as well as decisions for reimbursement and pricing assessments by the Transparency Commission.

However, both the HAS and industry stakeholders recognize that the current system is overly manual, complex, and time-intensive. Challenges such as limited resources, fragmented systems, and reliance on hospital clinicians who often complete PUT-RD forms outside of regular working hours have contributed to an average data completeness rate of just 65%, falling significantly short of HAS’s 90% target. These shortcomings undermine the evidential value of EAP data, which are critical for ensuring patient safety and supporting reimbursement decisions.

LEEM (Les Entreprises du Médicament), the industry’s representative, composed a white paper highlighting several barriers exacerbating these issues:

• Complex and inconsistent processes: Each industrial sponsor develops its own data capture system, resulting in duplication and confusion.
• Opaque agreements: Lack of transparency between industry and hospitals regarding compensation and data usage creates inefficiencies.
• Fragmented platforms: Limited interoperability discourages engagement and collaboration.
• Shortage of trained personnel: There is a notable lack of clinical research associates and data managers, particularly outside of major university hospitals.

The SNDS compiles health insurance claims for hospital stays and outpatient care across nearly the entire population, in addition to mortality data (including cause of death) and socio-demographic information. The database is updated continuously for outpatient claims and annually for hospital claims. All early access treatments provided by hospitals are traceable in the SNDS if they are reimbursed by National Health Insurance (i.e. when they are not provided free of charge). This opens the possibility of automatic extraction of key variables that are currently being requested in the PUT-RD, such as:

• Patient demographics (age, sex, region, comorbidities)
• Treatment data including previous or concomitant treatments, treatment duration, and estimated doses

While some variables (e.g. PROs or biological results) are unavailable, the SNDS still covers a substantial portion of what the HAS deems essential for contextual analysis. Furthermore, the SNDS can also be used to identify adverse events, as showcased in the work done by the public group EPI-PHARE. 

In short, the SNDS can automate the repetitive, descriptive part of the data collection process, freeing hospitals from a major administrative burden and improving completeness and comparability. A particularly promising use of SNDS integration is comparing patients receiving EAP drugs to those under standard care. Because SNDS records are population-wide, they allow for:

• Matching treated patients to comparable controls
• Evaluating real-world effectiveness relative to existing treatments
• Assessing healthcare resource use, hospitalization rates, and cost-effectiveness

Building on the feasibility study, the HAS has developed a prototype automated reporting tool designed to generate structured summaries of EAP utilization using SNDS data. These reports currently include key metrics such as:

• The number of patients treated per indication
• Patient demographics and comorbidities
• A description of the prescription setting

Future iterations of the tool could expand its capabilities to include dosage estimation, follow-up duration, treatment persistence, and interactive visualization dashboards for decision-makers. Automating these reports would significantly reduce feedback cycles, enabling near real-time surveillance and continuous optimization of EAP management.

Regulatory oversight could also be enhanced through shared dashboards across agencies, facilitating:

• Monitoring of ongoing EAPs
• Early detection of safety signals
• Data-driven recommendations for transitioning drugs to regular reimbursement pathways

The adoption of the SNDS and the reporting tool remains under discussion. Following a public consultation in early 2025, stakeholders are still awaiting the HAS’s final conclusions on its implementation.

EAPs remain a cornerstone of France’s policy for fostering innovation and ensuring timely access to treatments, particularly for drugs prior to marketing authorization. However, the data collection required to assess drug efficacy and safety has faced significant challenges, posing a substantial burden on clinicians and the industry alike.

The SNDS presents a transformative opportunity to streamline and scale RWD collection, replacing fragmented manual reporting with continuous, automated surveillance. By integrating clinical, administrative, and population-level data, France has the potential to lead Europe in real-world evidence–driven regulatory practices.

To realize this vision, industry and institutional stakeholders must align their efforts in governance, data infrastructure, and workforce training. Simplification, transparency, and automation are no longer optional; they are essential to preserving the credibility and efficiency of the EAP framework.