The integration of real-world evidence (RWE) into regulatory decision-making has been one of the most actively discussed shifts in pharmaceutical regulation over the past several years. Regulatory agencies have published frameworks and guidance. Industry has invested in real-world data infrastructure and analytical methods. And yet, despite the volume of activity around RWE, its actual use in drug approval decisions remains more limited, more conditional, and more technically demanding than much of the surrounding discourse suggests.
The gap between the aspiration for RWE in drug approvals and what agencies actually accept and under what conditions is where manufacturers and regulatory strategists most need clarity. This article provides a direct, technically grounded assessment of where RWE genuinely fits in the current approval landscape.
What Real-World Evidence Is and What It Is Not
Real-world data (RWD) refers to data on patient health and healthcare delivery collected outside conventional randomised clinical trials. Sources include electronic health records (EHRs), administrative claims databases, disease registries, and patient-generated data from digital health tools. Real-world evidence is the clinical evidence derived from analysis of that data.
This distinction matters because regulatory agencies evaluate RWE not as a category of data but as a set of analytical claims about clinical outcomes, claims that must be supported by methods capable of controlling for confounding, selection bias, informative censoring, and measurement error inherent in observational sources. The scientific question is never whether RWD exists; it is whether the analytical approach applied to that data produces evidence reliable enough to support a specific regulatory conclusion.
FDA Framework of 21st Century Cures Act and Current Guidance
FDA’s engagement with RWE for drug approvals is anchored in the 21st Century Cures Act (2016), which directed the FDA to develop a framework for evaluating RWE to support new indication approvals and post-approval study requirements.
FDA’s RWE Program guidance documents, updated through 2023–2025, establish conditions under which RWE may be considered for regulatory purposes. For RWE to support an efficacy claim, the FDA requires that the data source be fit for purpose: adequately capturing outcomes of interest, patient population, treatment exposure, and relevant confounders. Sponsors must pre-specify their analytical approach, demonstrate that the RWD source adequately captures clinically meaningful endpoints, and address potential biases through pre-specified sensitivity analyses. Post-hoc analytical decisions made after examining the data are not acceptable as the primary basis for a regulatory efficacy claim.
The FDA has approved a small number of supplemental indications based substantially on RWE, primarily in oncology, rare diseases, and paediatric populations, where randomised trial conduct is impractical. These approvals have not established a general RWE-for-efficacy pathway; they reflect the FDA’s case-by-case assessment that the totality of evidence supports the benefit-risk conclusion for that specific product and population.
EMA's Position (DARWIN EU and Current Regulatory Science)
EMA’s approach is articulated through its Regulatory Science Strategy to 2025 and operationalised through the Data Analysis and Real-World Interrogation Network (DARWIN EU), a federated network of real-world data sources across EU member states that EMA can query to support regulatory decisions.
EMA’s position is that RWE currently has its most established role in post-authorisation safety studies (PASS), benefit-risk assessments for pharmacovigilance, and evaluation of medicines in subpopulations under-represented in pivotal trials. For pre-approval efficacy claims, EMA’s standards remain high; the CHMP expects randomised controlled trial evidence as the primary basis for marketing authorisation, with RWE in a supporting role.
DARWIN EU is strategically significant because it gives EMA direct independent access to fit-for-purpose RWD without relying solely on sponsor-generated analyses. This means EMA can now independently query real-world data to validate or challenge sponsor claims, changing the evidentiary dynamic for any sponsor seeking to use RWE in an EMA submission.
Where RWE Is Currently Accepted
Across the FDA, EMA, MHRA, Health Canada, and TGA, RWE has established roles in specific, well-defined regulatory contexts.
External control arms: For rare diseases, paediatric conditions, or ultra-rare oncology indications where randomised placebo-controlled trials are impractical, agencies have accepted externally controlled studies using real-world patient data as the comparator arm. The RWD source must capture the same outcomes measured in the treated arm, the patient population must be demonstrably comparable, and the analytical approach must address confounding systematically.
Post-market safety commitments: PASS requirements attached to European marketing authorisations frequently mandate real-world pharmacoepidemiological studies. These are conditions of the marketing authorisation, not optional, and EMA expects pre-specified protocols, validated data sources, and independent statistical analysis.
Subgroup evidence and label expansion: For subpopulations where dedicated clinical trials are limited, paediatric patients, pregnant women, patients with renal or hepatic impairment, RWE from registries and EHR databases can supplement trial evidence to support labelling decisions. FDA has been progressively more receptive here, particularly for rare and paediatric indications.
Health technology assessment support: Post-approval RWE is critical for reimbursement decisions by HTA bodies, including NICE, HAS, and IQWiG. High-quality post-market RWE is a commercial asset, not only a regulatory deliverable.
What Agencies Will Not Accept
RWE cannot serve as the primary efficacy basis for a first-indication approval where randomised trial evidence is feasible. The absence of RCT evidence is not itself a justification; the justification must be the impracticability of conducting the trial, not commercial or logistical inconvenience.
Post-hoc generated RWE, where the analytical approach is specified after examining the data, will not support efficacy or safety claims. Pre-specification is non-negotiable. Poorly characterised data sources, where the completeness, accuracy, and representativeness of the underlying RWD are undocumented, will not satisfy fit-for-purpose requirements regardless of analytical sophistication.
Conclusion
Real-world evidence has a genuine and growing role in regulatory decision-making, but it operates within methodological standards that are demanding and agency expectations that are specific. The strongest current uses external control arms for rare diseases, post-authorisation safety studies, subpopulation evidence, and HTA support are well-defined. The aspiration to use RWE as a broad substitute for randomised trial evidence in efficacy-based approvals remains, for most indications, unrealised and unlikely to be accepted under current agency frameworks. Manufacturers who engage RWE strategically, building pre-specified, fit-for-purpose programmes aligned with what agencies actually accept extract genuine regulatory value from real-world data. Those who approach it as a shortcut to avoiding clinical trial investment will consistently find that agencies disagree.
How DDReg can help?
DDReg supports pharmaceutical and biopharmaceutical companies in designing and implementing RWE strategies aligned with current FDA, EMA, MHRA, Health Canada, and TGA regulatory expectations from RWD source assessment and study protocol pre-specification through regulatory submission integration and post-authorisation PASS programme management.
Read more from our experts here: COFEPRIS Cosmetic Approval in Mexico: What’s Right for Your Product?