The Regulatory Shift Toward Living Evidence
The FDA and EMA are converging on continuous, real-time evidence requirements. From the single-trial policy to DARWIN EU, here's what's changing and why it matters for evidence synthesis.
The evidence landscape is being rewritten
In the span of four months — from December 2025 to March 2026 — the FDA has enacted three regulatory changes that collectively signal a fundamental shift in how drug evidence is generated, evaluated, and maintained. These aren't incremental updates. They represent a new regulatory philosophy: evidence must be continuous, not static.
For researchers, institutions, and pharmaceutical companies, this shift has immediate implications for how evidence synthesis is conducted. Here's what has changed.
The FDA's three moves
1. De-identified real-world evidence accepted (December 2025)
On December 18, 2025, the FDA announced it would accept real-world evidence without requiring identifiable individual patient data in regulatory submissions. Previously, the requirement for private patient-level data made most large healthcare databases impractical to use. Under the new policy, de-identified data from cancer registries, hospital systems, insurance claims databases, and electronic health record networks can be submitted on a case-by-case basis.
The scale of change this enables is significant. Since 2016, only 35 drugs, biologics, or vaccines had included real-world evidence in their applications — compared to over 250 medical device authorizations. This policy removes the primary barrier that created that disparity.
2. Single pivotal trial as default (February 2026)
In February 2026, FDA Commissioner Martin Makary and vaccine chief Vinay Prasad wrote in the New England Journal of Medicine that advances in biology, trial methodology, and statistical science justify ending the longstanding "two-trial dogma" for new drug applications. The FDA will now accept a single adequate and well-controlled clinical trial combined with confirmatory evidence as the default basis for approval.
This change intensifies the importance of post-market evidence. With fewer pre-market trials, the burden of ongoing evidence generation — including real-world evidence, post-approval studies, and continuous safety monitoring — increases substantially.
3. Bayesian methodology guidance (January 2026)
The FDA issued draft guidance on the use of Bayesian methodology in clinical trials for drugs and biological products. Bayesian approaches allow trial designs that adapt as evidence accumulates — a continuous learning process that aligns with the broader move toward living evidence. Traditional frequentist methods typically require four to five years per trial; Bayesian designs can reduce both time and patient exposure to ineffective treatments.
FDORA's accelerated approval reform
The consequences of failing to maintain post-market evidence are already visible. Under the Food and Drug Omnibus Reform Act (FDORA), the FDA withdrew eight cancer drug accelerated approvals in 2024 alone. Pepaxto (melphalan flufenamide) was the first withdrawal under FDORA's expedited procedures, completed in approximately seven months — compared to 30 months for the pre-FDORA Makena withdrawal.
The median time to withdrawal has decreased from 9.5 years to 3.2 years. The proportion of accelerated approvals with confirmatory trials already underway at approval has risen from 63% to 85%. The message is clear: sponsors must have continuous evidence plans or face rapid consequences.
The FDA's evidence infrastructure
The FDA is not just demanding continuous evidence — it's building the infrastructure to evaluate it. The Sentinel Real-World Evidence Data Enterprise (RWE-DE) now covers more than 25 million lives through linked electronic health records and insurance claims data. Since its inception, the system has completed 784 analyses resulting in 12 drug labeling changes, advisory committee presentations, and drug safety communications.
The agency also launched FDA-RWE-ACCELERATE, an agency-wide initiative running through fiscal year 2027 that provides sponsors with pre-protocol meetings to discuss real-world evidence approaches for new labeling claims and post-approval requirements.
The European side: DARWIN EU and mandatory RWE
The shift is not limited to the United States. The European Medicines Agency's DARWIN EU network has expanded to approximately 40 data partners across 16 countries, covering roughly 160 million patients. The network now delivers approximately 100 studies per year, with a median turnaround of four months from protocol approval to results.
More consequentially, EU Implementing Regulation 2025/1466, which took effect in February 2026, mandates continuous EudraVigilance monitoring for prompt signal detection, electronic registration of post-authorization safety studies before data collection, and integration of real-world evidence into pharmacovigilance systems. Marketing Authorization Holders must now systematically incorporate real-world data throughout their signal management processes.
Living evidence in pharmacovigilance
Drug safety monitoring teams are already adopting living evidence reviews — systematic reviews that continuously update as new data emerge on defined schedules or when safety triggers occur. These reviews maintain the scientific rigor of traditional systematic reviews while allowing pharmacovigilance teams to keep safety summaries, signal assessments, and periodic safety update reports current without rebuilding evidence bases from scratch.
According to Deloitte's benchmarking data, 90% of biopharmaceutical executives say their organizations are attempting to leverage real-world evidence across the product lifecycle.
The reporting standard exists
The infrastructure for standardized living reviews is also maturing. PRISMA-LSR, published in The BMJ in November 2024, provides a reporting extension specifically for living systematic reviews — including version control documentation, search update protocols, and criteria for transitioning reviews out of living mode.
What this means
The regulatory apparatus — FDA, EMA, and national agencies — is converging on a single expectation: evidence must stay current. Static systematic reviews that take 30 weeks to produce and are never updated will not meet the evidentiary standards that regulators are building toward.
At Mapped Technologies, we see this convergence as the defining challenge of the next decade of medical research infrastructure. Our current product, mapped, covers the complete systematic review workflow. Our next product is designed for exactly this shift: living evidence synthesis that is continuously updated, fully auditable, and built to meet regulatory-grade requirements.
The infrastructure for continuous evidence doesn't exist yet at production scale. We're building it.