Real World Data and Real World Evidence to Support Regulatory Decision Making for Drug and Biological Products – FDA Draft Guidance

Use of Real-World Data (RWD) and Real-World Evidence (RWE) in Regulatory Decisions.

The FDA issued this draft guidance as part of its Real-World Evidence (RWE) Program to help fulfill requirements under section 505F of the FD&C Act. The guidance explains how RWE may support approval of new indications for already-approved drugs under section 505(c) of the FD&C Act or help meet post-approval study obligations.

Regulatory Considerations

1. Applicability of 21 CFR Part 312 
2. Part 312 outlines the regulatory framework for Investigational New Drug (IND) applications. FDA recognizes that RWD can be used effectively in interventional studies—such as identifying potential participants for randomized trials, supporting endpoint assessment, or serving as an external comparator arm in historically controlled trials.
3. Non-interventional (observational) studies rely on data derived from the routine clinical use of marketed drugs, guided by providers’ medical judgment rather than study-specific treatment assignments.

Regulatory Considerations for Non-Interventional Studies

1. Transparency in Data Collection and Analysis 

• Sponsors should consult FDA early when developing a non-interventional study for regulatory submission. Draft protocols and statistical analysis plans (SAPs) should be submitted for review before finalization and before any study analyses begin.

• FDA must be assured that study design decisions—including data source selection and analytic plans—are made prior to assessing study outcomes. All protocol and SAP updates must be dated and justified. Sponsors should:

• describe all data sources accessed during study design; 
• explain why data sources were selected or excluded; 
• maintain audit trails showing all dataset access and analyses performed; 
• document all feasibility assessments and exploratory analyses; 
• describe source and study populations and explain differences that could influence results.

• Protocols should be posted publicly, such as on ClinicalTrials.gov or the ENCePP registry.

2. Access to RWD 

• Sponsors must discuss RWD access expectations with FDA early in study planning. They must also ensure they can submit patient-level data for all RWD used in a marketing application, as required under 21 CFR 314.50 and 601.2.
• If RWD is owned by third parties, contractual agreements should guarantee FDA access to all necessary patient-level and source data. All datasets and programming code submitted must be complete, clearly documented, and sufficiently annotated to allow FDA to replicate analyses.

3. Study Monitoring 

• For purely observational studies, monitoring should focus on ensuring the reliability of RWD—from data extraction through final reporting. If the study requires additional protocol-driven procedures, monitoring must also confirm compliance with human subject protection requirements.
• FDA encourages risk-based quality management, emphasizing oversight of processes most critical to study quality and participant protection.

4. Safety Reporting 

• Sponsors submitting NDAs or BLAs remain responsible for post-marketing safety reporting. Because observational studies use data from routine clinical practice, any adverse events identified during the study that meet reporting requirements must be submitted to FDA.
• If only part of a large RWD source is used for an analytic dataset, sponsors are not expected to search the entire database for additional adverse events unrelated to the labeling objective.

5. Additional Sponsor Responsibilities 

For marketing applications containing non-interventional studies, electronic systems used to manage study data must comply with 21 CFR Part 11.

Sponsors must take responsibility for:

• selecting qualified researchers; 
• ensuring conduct consistent with the final protocol and SAP; 
• maintaining complete and accurate study records; 
• ensuring FDA access to source data and records; 
• implementing appropriate monitoring procedures.

Sponsors must maintain a log of all researchers substantially involved in the study, including their names, roles, affiliations, and qualifications. If third parties (e.g., data vendors, CROs) are involved, sponsors should document their roles and responsibilities and provide this information to FDA upon request. Sponsors remain accountable unless responsibilities have been formally transferred to a contract research organization.