Bhargav Mantha, Sundeep Karnik and Anna Sato contributed to this article.

Key takeaways

• Start with the decision, not the available data. Decision-grade real-world evidence (RWE) is most valuable when it can influence high-stakes regulatory, clinical, payer or product decisions—from label expansion and post-market commitments to payer coverage and reimbursement, product differentiation, AI-enabled updates and market access strategy. Defining the use case first clarifies what evidence is needed to credibly shape action.
• Fit-for-purpose RWE depends on context as much as data access. Medtech teams need to trace device exposure, connect the relevant clinical and economic context and use methods that regulators, payers and clinicians can review and trust.
• The greater advantage comes from making RWE repeatable. Reusing cohorts, definitions, methods, quality checks and governance lessons can turn one-off studies into a managed capability that makes each new use case faster, more consistent and more credible.

As real-world data (RWD) becomes more available in medtech, the question for leaders is no longer whether it can be used, but whether it can be turned into evidence credible enough to shape regulatory, access, clinical and product decisions. Can an indication expansion be supported? Is technology performing as expected in broader populations? What will payers and providers expect? How should an AI-enabled product be monitored as practice patterns evolve?

Without a repeatable foundation for generating decision-grade RWE, medtech teams will continue rebuilding evidence pathways for each new need, leading to longer timelines, inconsistent standards, delayed label expansion, unresolved payer questions, slower adoption, weaker product differentiation and limited cumulative value across the enterprise.

The next RWE advantage in medtech will belong to first movers that build connected, repeatable evidence-generation systems—converting fragmented real-world signals into decision-ready evidence faster, with reusable assets that strengthen each successive regulator, payer, product and adoption decision.

Why does medtech need its own RWE model?

Medtech can accelerate RWE maturity by adapting pharma’s discipline in protocolized design, governance, reusable cohorts, transparent assumptions and evidence operations—but not by copying its model. Device exposure, iteration and real-world use are more variable and often require integration across procedure codes, product identifiers, unique device identifiers, implant logs, supply-chain records, operator notes and software versions.

Those exposure challenges sit inside a more variable clinical and operational context. Product iteration, operator experience, procedural technique, care setting (patient selection and site-level workflow), learning curves, follow-up intensity, device generation, component-level variation and software changes can all shape device performance. Peer-reviewed medical device RWE literature has highlighted similar challenges, including rapid design iteration, the role of operator expertise and difficulty implementing blinding or placebo controls in many device studies.

RWE value should be created through a disciplined, stepwise planning flow before analysis begins. Teams should frame the decision, define the required confidence level, test whether data sources can be linked and interpreted in context, specify exposure and endpoints and document assumptions and governance. This makes the evidence credible, reviewable, reusable and responsive to product life cycle needs.

What can recent RWE precedents teach medtech leaders?

The FDA’s updated medical device RWE guidance provides further clarity and requirements for medtech sponsors by making clear that RWE must be fit for the specific regulatory decision. The path forward requires earlier and more disciplined planning around relevance, reliability, data provenance, linkage, exposure and endpoint definitions, bias control, statistical methods, governance and documentation, while also expanding practical pathways for RWE to support device decisions across randomized, nonrandomized, single-arm, observational and hybrid study designs.

The FDA’s current set of 73 RWE examples demonstrates that this is no longer theoretical: The FDA’s Center for Devices and Radiological Health (CDRH) has already authorized device decisions using RWE across 510(k), de novo, humanitarian device exemption (HDE), premarket approval (PMA) and PMA supplement pathways. In these examples, RWE served as a primary source of clinical evidence, supported multisource linkage and hybrid designs and helped validate software, AI and machine learning-enabled device functions. For medtech sponsors, the implication is clear: RWE can be a strategic evidence asset when it is built for a specific decision and meets FDA expectations for decision relevance, data reliability and review-ready documentation.

The FDA’s public RWE examples show that medtech sponsors are already using fit-for-purpose RWE to support high-stakes device decisions. Examples include clinical trial pathway transitions, retrospective cohort studies using medical record data, linked data from electronic health records (EHRs), claims and mortality sources for product evolution and multisource evidence packages for indication expansion. Across these use cases, data sources, exposure, endpoints, follow-up, uncertainty and documentation are defined up front.

Across these examples, the common thread is a clear use case, fit-for-purpose sources, traceable device exposure, documented linkage, credible methods and a package stakeholders can review. The strategic advantage comes from preserving those elements for reuse rather than rebuilding them one study at a time.

How can medtech build RWE as a repeatable capability?

Medtech RWE maturity depends on moving from a sequence of studies to a managed capability that can keep pace with connected devices, software updates, evolving clinical practice and changing stakeholder expectations. That capability connects priority use cases, fit-for-purpose sources, validated exposure definitions, credible methods, traceable documentation and reusable assets that can be refreshed as products and priorities change.

Used responsibly, agentic AI can help teams identify emerging needs, assess source fitness, test linkage and endpoint feasibility, draft traceable documentation and refresh reusable assets as devices, coding, data sources and stakeholder expectations change. The goal is not autonomous regulatory decision-making. AI should help prepare, connect and document the evidence workflow; people should make the scientific, clinical and regulatory judgments. AI can improve consistency and traceability while keeping people accountable for the judgment.

For AI-enabled devices, life cycle evidence management is closely tied to product change. The FDA’s 2025 guidance on predetermined change control plans describes how planned modifications, validation methodology and impact assessment can be included in a marketing submission. The International Medical Device Regulators Forum’s (IMDRF) 2025 Good Machine Learning Practice principles emphasize life cycle application, data representativeness, traceability, reproducibility and monitoring. In Europe, Article 72 of the EU AI Act requires post-market monitoring systems for high-risk AI systems that actively collect, document and analyze performance and compliance data throughout the system’s lifetime.

5 capabilities that make connected evidence systems work

A practical operating model brings together five connected capabilities under common standards and reusable ways of working. Together, they can also serve as a practical diagnostic for whether an organization is ready to generate decision-grade RWE repeatedly. The aim is not to centralize every study. That shared foundation helps regulatory, clinical, medical, quality and data teams work with health economics and outcomes research colleagues across priority use cases and the product life cycle. Below are those capabilities and questions leaders must answer to put them into practice.

Where should medtech sponsors start with RWE? Where evidence can change a decision

Medtech sponsors should start with two or three high-value life cycle decisions where external scrutiny, budget pressure or growth ambition is already visible—such as label expansion, payer pushback, premium differentiation, post marketing commitments, AI-enabled updates, safety surveillance or unresolved evidence gaps that slow adoption.

For each priority decision, leaders can move through three gates.

• Define the decision. What must the evidence support, who needs to trust the result and what level of uncertainty is acceptable?
• Pressure-test feasibility. Are the available data sources, linkages, exposure definitions, endpoints, comparators and methods strong enough to support the decision?
• Productize what works. Capture the cohorts, code lists, endpoint definitions, quality checks, templates, analytic workflows and governance lessons, so the next evidence question starts from a stronger foundation.

The value compounds. Each well-designed RWE effort should solve an immediate need while making the next one faster, more consistent and more credible across products, indications, markets and stakeholders.

For medtech leaders, the value will accrue first to those who build managed, fit-for-purpose RWE capabilities that stay connected to clinical context and trusted enough to shape decisions, turning each study into a reusable evidence asset that compounds across products, indications and markets.

Note: This document is for discussion only and is not intended as regulatory advice. Sponsors should consult applicable laws, regulations, guidance and agency feedback for specific programs.