FDA closed out 2025 with a meaningful update for device sponsors: the final guidance, “Use of Real-World Evidence (RWE) to Support Regulatory Decision-Making for Medical Devices,” issued December 18, 2025, which supersedes the prior 2017 version.

At QSN, we see this as more than a “nice-to-have” document. It’s a clearer playbook for when FDA believes real-world data (RWD) can be leveraged to generate regulatory-grade evidence, as well as what you’ll need to show to make it credible in a submission.

First, the basics (as FDA defines them)

• RWD: routinely collected data relating to patient health status and/or healthcare delivery (EHRs, claims, registries, patient-generated data, device-generated data, etc.).
• RWE: the clinical evidence about a device’s use and potential benefits/risks derived from analysis of RWD.

FDA’s bottom line stays the same: RWE can support regulatory decisions under the right conditions, and FDA is clarifying when RWD quality is sufficient to meet those standards.

The “QSN lens”: Relevance and Reliability

FDA repeatedly comes back to two core questions:

1) Relevance: Does the data actually answer the regulatory question?

FDA will look for whether the RWD:

• captures the right population, device exposure, comparators (if applicable), and outcomes of interest
• contains enough clinical detail to support analysis (and interpretation) for your specific claim
• supports appropriate confounding adjustment (where needed) and has a pre-defined common data model / dictionary
• can identify the device with adequate granularity (often meaning UDI/model/serial-level traceability)

Practical takeaway: if you can’t clearly map your intended use/indication, endpoints, and device identification into the dataset, you’re likely looking at “RWD for context,” not “RWD for clearance/approval.”

2) Reliability: Can FDA trust how the data were captured and controlled?

Reliability hinges on:

• Data accrual: how the data were collected (definitions, time windows, site readiness, capture methods, completeness)
• Data assurance / QC: training, monitoring/auditing strategy, consistency checks, validation, and an overall “quality system mindset” for the data pipeline

Practical takeaway: treat your RWD source like a manufacturing process. If you can’t describe controls, validation, and change management, FDA will have a hard time relying on it.

Where FDA expects RWE to show up across the product lifecycle

FDA explicitly describes RWE as potentially useful for decisions across the Total Product Life Cycle, including:

• supporting or supplementing premarket submissions (where appropriate)
• label expansion / new indications
• postmarket surveillance, PAS/522 strategies (sometimes enabling reduced premarket burden)
• signal evaluation and public health surveillance

A key operational topic sponsors often miss: IDE implications

FDA clarifies a real-world “line in the sand”:

• If a legally marketed device is used in the normal course of medical practice, RWD collection often won’t require an IDE.
• If your data collection influences treatment decisions or creates prospective research-like behavior to determine safety/effectiveness (e.g., a registry designed for a new intended use with prescribed follow-up requirements), an IDE may be required.

Practical takeaway: before you invest in a prospective RWD study, pressure-test the protocol against IDE triggers early (and don’t be afraid to use FDA’s pre-sub process to align).

What’s “new” feeling about this final 2025 update?

Beyond consolidating and modernizing RWE expectations, FDA also signaled a meaningful reduction in friction: FDA stated it removed a major limitation by indicating it will accept certain RWE without always requiring identifiable individual patient data to be submitted in device (and drug) application reviews.

Why that matters: it can lower operational and privacy barriers in some contexts, but it does not remove your obligation to demonstrate that the evidence is reliable, auditable, and fit-for-purpose.

QSN recommended next steps (a fast internal checklist)

If you’re considering RWE for a device submission in 2026:

1. Define the regulatory question (decision type + claim + endpoints)
2. Confirm device traceability in the dataset (UDI/model/serial where needed)
3. Build a protocol + analysis plan before access (even for retrospective work)
4. Document data QC/assurance like a quality system (roles, training, monitoring, auditability)
5. Use Pre-Sub strategically to de-risk acceptance

If you’re navigating RWE strategy, IDE considerations, or FDA alignment for 2026 submissions, QSN can help you build evidence FDA is prepared to rely on.