Real-World Evidence (RWE) for Medical Devices (2026): FDA Guidance, Data Sources, Regulatory Uses & Implementation Guide

What Is Real-World Evidence?

Real-world evidence (RWE) is clinical evidence about the usage, benefits, or risks of a medical product derived from the analysis of real-world data (RWD). RWD is data relating to patient health status and/or the delivery of health care routinely collected from a variety of sources outside traditional clinical trials.

The distinction is important: RWD is the data; RWE is the evidence generated from analyzing that data.

FDA Definitions (December 2025 Final Guidance)

Term	Definition
Real-World Data (RWD)	Data relating to patient health status and/or the delivery of health care routinely collected from a variety of sources
Real-World Evidence (RWE)	The clinical evidence regarding the usage, and potential benefits or risks, of a medical product derived from analysis of RWD

Sources of Real-World Data

RWD Source	Description	Examples
Electronic Health Records (EHRs)	Digital records of patient encounters in clinical settings	Epic, Cerner, VA health system data
Medical Device Registries	Organized systems collecting uniform clinical data on device outcomes	National registries, specialty society registries
Administrative Claims Data	Billing and claims records from insurers and healthcare systems	Medicare claims, commercial insurance databases
Patient-Generated Health Data	Data collected by patients through digital health technologies	Wearable devices, home monitoring, patient-reported outcomes
Public Health Databases	Government-maintained health surveillance data	State health department data, vital statistics
Digital Health Technologies	Data from connected medical devices and software	Continuous glucose monitors, remote cardiac monitors

The December 2025 FDA Final Guidance: A Landmark Update

On December 18, 2025, the FDA published a finalized revision to its 2017 guidance document, "Use of Real-World Evidence to Support Regulatory Decision-Making for Medical Devices." This update represents the most significant policy change for RWE in medical device regulation in nearly a decade.

What Changed from the 2017 Guidance

The 2017 guidance required sponsors to be able to provide identifiable, individual-level participant data derived from RWD sources upon FDA request. This created a significant barrier — many large healthcare databases (claims databases, de-identified registries) restrict access to identifiable data, making them effectively unusable for regulatory submissions.

The December 2025 update removes this barrier. FDA now accepts RWE generated from appropriately analyzed de-identified and aggregate data from large healthcare datasets, including registries and claims databases, without requiring access to individually identifiable source data.

Key Policy Shifts

2017 Guidance	December 2025 Update
Sponsors expected to provide identifiable participant-level data upon request	RWE evaluated case-by-case; identifiable data not always required
Large de-identified databases were effectively excluded	Large de-identified and aggregate datasets now accessible for regulatory use
Focus was on verifying individual source data	Focus shifted to scientific soundness and data fitness for purpose
Limited examples of RWE use in practice	FDA published detailed examples catalog (April 2026) covering FY2020–2025

Transition Period

The guidance includes a 60-day transition period. Beginning February 17, 2026, FDA anticipates that sponsors will include the newly recommended information in their submissions. However, FDA will review such information if submitted at any time.

How FDA Evaluates RWE: Relevance and Reliability

The FDA evaluates RWE submissions based on two core criteria: relevance and reliability.

Relevance

Relevance considers whether the data is appropriate for answering the specific regulatory question:

• Sufficiency in detail: Does the data capture the clinical variables, device identifiers, and outcomes needed?
• Representativeness: Does the study population represent the intended use population?
• Appropriate comparator: Is there a suitable control group or comparator?
• Adequacy of follow-up: Is the follow-up duration sufficient to capture relevant outcomes?
• Temporal relevance: Is the data current enough to reflect current clinical practice?

Reliability

Reliability encompasses the integrity and quality of the data:

• Data completeness: Are critical data fields populated consistently?
• Data accuracy: Are there validation mechanisms to ensure data quality?
• Consistency: Are data collection methods uniform across sites and time?
• Provenance: Is the origin and chain of custody of the data well-documented?
• Quality assurance: Are there audit trails and quality control processes?

Regulatory Applications of RWE

RWE can be used across the total product lifecycle of a medical device. The FDA identifies multiple applications:

Premarket Applications

Application	Description	Submission Type
Primary clinical evidence	RWE as the main evidence supporting marketing authorization	510(k), De Novo, PMA
Historical control	RWE used to construct a historical control group for single-arm studies	PMA, De Novo, Breakthrough
Concurrent control	RWE providing a real-world comparator arm	Clinical investigations
Hypothesis generation	RWE identifying signals that inform clinical study design	Pre-submission
Supplemental evidence	RWE supporting or strengthening traditional clinical evidence	Any submission type
Labeling expansion	RWE supporting new indications or expanded patient populations	PMA supplement, 510(k)

Postmarket Applications

Application	Description
Post-approval studies	Fulfilling post-approval study requirements imposed as conditions of approval
Post-market surveillance	Ongoing monitoring of device performance in real-world settings
Labeling updates	Supporting safety and effectiveness claims for labeling changes
Signal detection	Identifying potential safety signals from real-world use data
Recall support	Informing recall decisions and scope based on real-world performance data

FDA's Published RWE Examples (FY2020–2025)

In April 2026, the FDA published "Examples of Real-World Evidence Used in Medical Device Regulatory Decisions (FY2020–2025)", a comprehensive catalog of 73 real-world examples of marketing authorizations. Key examples include:

Device Approvals Using Registry Data

Several devices received marketing authorization based in part on data from national device registries. These registries collected uniform clinical data across multiple sites, providing robust evidence of device performance in real clinical practice.

Claims Database Analyses

Retrospective analyses of insurance claims data were used to support both premarket submissions and postmarket surveillance activities. These analyses demonstrated device outcomes in large, diverse patient populations over extended follow-up periods.

EHR-Derived Evidence

Electronic health record data was used to characterize device performance in specific patient subgroups and to support labeling expansions based on real-world clinical experience.

Hybrid Study Designs

Some submissions combined traditional clinical trial data with RWE, creating hybrid evidence packages that leveraged the rigor of prospective trials with the generalizability of real-world data.

Practical Guide: Incorporating RWE into Your Regulatory Strategy

Step 1: Identify the Regulatory Question

Clearly define what the RWE needs to demonstrate:

• Is it supporting device safety, effectiveness, or both?
• Is it the primary evidence or supplemental?
• What is the specific regulatory pathway (510(k), De Novo, PMA)?

Step 2: Assess Data Source Suitability

Evaluate potential RWD sources against FDA's relevance and reliability criteria:

Criterion	Key Questions
Data captures device and outcomes?	Does the source collect the device identifiers, procedural details, and clinical outcomes you need?
Population is representative?	Does the data cover your intended patient population (age, comorbidities, practice settings)?
Sample size is adequate?	Is there sufficient data to support statistical analyses?
Follow-up is adequate?	Does the data capture outcomes over a clinically meaningful time period?
Data quality is verifiable?	Are there quality assurance processes, audit trails, and validation mechanisms?

Step 3: Design the Study Protocol

Develop a protocol that specifies:

• Study design (retrospective cohort, case-control, cross-sectional)
• Inclusion and exclusion criteria
• Primary and secondary endpoints
• Statistical analysis plan
• Bias mitigation strategies
• Sensitivity analyses

Step 4: Engage FDA Early

The FDA strongly recommends early engagement through:

• Pre-submission meetings (Q-submissions) to discuss your RWE approach
• FDA town halls and public workshops on RWE topics
• CDRH's Office of Clinical Evidence and Analysis for specific questions

Step 5: Document Data Provenance and Quality

Prepare a comprehensive data dossier including:

• Description of the RWD source and data collection methods
• Data governance and quality assurance processes
• Completeness analysis and missing data handling
• Validation study results (if available)
• Linkage methods (if combining multiple data sources)

Step 6: Prepare the Submission

Include in your regulatory submission:

• Study protocol and statistical analysis plan
• Complete results with appropriate statistical methods
• Sensitivity analyses and bias assessment
• Comparison to traditional clinical evidence (if applicable)
• Discussion of limitations

RWE vs. Traditional Clinical Evidence

Aspect	Traditional Clinical Trials	Real-World Evidence
Setting	Controlled experimental environment	Routine clinical practice
Population	Selected per strict inclusion/exclusion	Broader, more representative
Sample size	Often limited by enrollment	Can be very large (thousands to millions)
Follow-up	Fixed protocol-defined periods	Variable, potentially years
Cost	High ($millions to $tens of millions)	Lower (leveraging existing data)
Timeline	Years to complete	Months to analyze
Bias	Minimized by randomization and blinding	Potential for confounding and selection bias
Generalizability	Limited to trial population	Higher — reflects real clinical practice
Control	Active or placebo control	Historical control, concurrent registry, or claims-based
FDA acceptance	Gold standard	Increasingly accepted; case-by-case evaluation

RWE and Digital Health Technologies

The intersection of RWE and digital health is creating new opportunities:

Connected Devices and Continuous Monitoring

Devices like continuous glucose monitors, cardiac implantable electronic devices, and remote patient monitoring systems generate vast quantities of real-world performance data. This data can be used for:

• Post-market performance monitoring
• Supporting labeling updates based on real-world outcomes
• Generating evidence for new indications

AI/ML-Enabled Devices

For AI/ML-enabled medical devices, RWE is particularly valuable for:

• Monitoring algorithm performance in diverse clinical settings
• Detecting algorithmic drift as models encounter new data patterns
• Supporting Predetermined Change Control Plans (PCCPs)

FDA's NEST Network

The National Evaluation System for health Technology (NEST) continues to develop as a network of data sources that FDA can reference for regulatory decision-making. NEST aims to generate better evidence faster and at lower cost by coordinating real-world data across multiple healthcare systems.

International Perspective: EU and RWE

The European Medicines Agency (EMA) and EU regulatory framework are also expanding the use of RWE:

• The EU MDR allows for clinical evidence to include data from post-market surveillance, clinical follow-up, and registries
• EMA's Big Data Steering Group has published recommendations on using real-world data for regulatory purposes
• EU Drug Development Innovation Network (EU-DRAIN) initiatives are exploring RWE methodologies

Manufacturers pursuing both US and EU market access should consider designing RWE strategies that satisfy both FDA and EU requirements.

Key Considerations for 2026

Opportunities

1. De-identified data now accepted — The biggest barrier to using large healthcare databases has been removed
2. FDA actively encouraging RWE use — Multiple FDA communications in 2025–2026 signal strong institutional support
3. Published examples provide precedent — The FY2020–2025 examples catalog gives sponsors concrete reference cases
4. NEST network expanding — More data sources and better linkage methods are becoming available
5. Cost and timeline advantages — RWE can reduce the cost and time of evidence generation compared to traditional trials

Challenges

1. Data quality varies — Not all RWD sources meet FDA's relevance and reliability standards
2. Case-by-case evaluation — There is no guaranteed acceptance; each submission is individually assessed
3. Confounding and bias — Observational data requires careful study design and statistical methods to address biases
4. Regulatory precedent is still developing — While growing, the body of accepted RWE submissions remains limited compared to traditional evidence
5. Data access and governance — Securing access to large healthcare databases may require complex data use agreements

FAQ

Can RWE completely replace a clinical trial for FDA device approval?

In certain circumstances, yes. The FDA states that RWE may be used to supplement or, in some cases, partially replace traditional clinical evidence when adequately justified. However, whether RWE alone is sufficient depends on the device, the risk classification, the regulatory question, and the quality of the available data.

What types of submissions can include RWE?

RWE may be included in 510(k) submissions, De Novo classification requests, PMA applications, PMA supplements, HUD/HDE applications, and post-approval study reports.

Does the December 2025 guidance apply to drugs and biologics?

No. The December 2025 update applies to medical devices only. FDA has indicated it intends to consider updating parallel RWE guidance for drugs and biologics in the future.

How do I know if my RWD source meets FDA standards?

Evaluate your data source against FDA's relevance and reliability criteria. Consider engaging FDA through a Pre-submission meeting (Q-submission) to discuss the suitability of your proposed RWD source before committing resources to a full analysis.

What is NEST and how does it relate to RWE?

The National Evaluation System for health Technology (NEST) is a network of health data partners coordinated by the FDA's CDRH. NEST provides a framework for accessing and analyzing real-world data from multiple sources to support regulatory decision-making.

Can I use RWE from non-US data sources?

Yes, but the FDA will evaluate whether the non-US data is relevant to the US patient population and clinical practice. Factors include differences in clinical practice, patient demographics, device versions, and healthcare system structure.

Key Takeaways

• The December 2025 FDA final guidance is the most significant RWE policy update in nearly a decade — it removes the requirement for identifiable patient-level data, opening the door to large de-identified databases
• RWE can serve as primary or supplemental evidence across premarket submissions (510(k), De Novo, PMA) and postmarket activities
• Relevance and reliability are the two core criteria FDA uses to evaluate RWE submissions
• Early FDA engagement through pre-submission meetings is strongly recommended before committing to an RWE-based strategy
• Published FDA examples (FY2020–2025) provide concrete precedents for how RWE has been successfully used in device regulatory decisions
• Digital health technologies and connected devices are expanding the sources and volume of real-world data available for regulatory purposes
• RWE does not replace all clinical trials — it is a complementary tool that, when properly designed and executed, can strengthen regulatory submissions and reduce evidence-generation costs