EMA and MHRA raise the design bar for real-world external controls

What changed

Regulatory teams should no longer treat a real-world external control as a dataset to add after trial results are known. EMA and MHRA are moving toward frameworks that require the comparison to be designed and documented in advance.

EMA’s final reflection paper sets methods for using real-world data in non-interventional studies for regulatory purposes. EMA has also adopted a concept paper for future guidance on external controls derived from real-world data, other trials or other sources. The planned paper will address when external controls could support pivotal or supplementary evidence. (ema.europa.eu)

MHRA’s more detailed external-control document remains a draft. Its consultation closed on July 14, 2025, and the agency’s current page says responses will inform final guidance. (gov.uk)

MHRA also relaunched its joint RWE Scientific Dialogue with NICE on May 20, 2026. The service uses workshops to discuss cross-cutting study designs, data sources and analytical methods across regulatory and health technology assessment decisions. (gov.uk)

Why it matters

Neither regulator presents external controls as a routine substitute for randomisation. EMA calls randomised controlled trials the evidentiary gold standard. The MHRA draft says a fully powered randomised trial should be used where possible and prefers a randomised trial augmented with external data over a single-arm trial supported only by an external control. (ema.europa.eu)

Where sponsors propose an external control, the MHRA draft expects a trial-quality protocol. It calls for robust justification of the data source and comparator construction, prespecified endpoints and analyses, and an explicit plan to identify and address bias. (assets.publishing.service.gov.uk)

Teams must also align populations, treatments, endpoints and timing. The estimand—the treatment effect the study is designed to estimate—should match across the trial and external control. Time zero and follow-up should be comparable, and missingness must not make the external dataset unsuitable. (assets.publishing.service.gov.uk)

The practical response is an auditable evidence-governance plan. It should cover data provenance and fitness for purpose, comparator eligibility, endpoint and estimand alignment, timing, prespecified analyses, missing data, confounding and sensitivity analyses. This is a prudent planning standard, not a new universal legal obligation.

Who should pay attention

Medicine sponsors considering single-arm trials, hybrid randomised designs or real-world external controls should involve regulatory affairs, biostatistics and epidemiology before protocol finalisation.

Data custodians and research partners should expect more scrutiny of provenance, completeness, coding, endpoint availability and whether the source is fit for the intended comparison.

UK medtech teams can also propose cross-cutting RWE topics to the MHRA-NICE dialogue, which now accepts medical-device applications and can include Approved Bodies. The dialogue is not product-specific regulatory advice. (gov.uk)

What to watch next

EMA’s concept paper schedules CHMP discussion of a draft external-controls reflection paper for the fourth quarter of 2026, followed by a three-month consultation. Final adoption is expected in the second quarter of 2027. (ema.europa.eu)

Watch for MHRA’s consultation-response summary and final external-control guideline. As of August 11, 2026, the official material still labels the document as draft. (gov.uk)

The current MHRA-NICE dialogue application window closes at 11:59 p.m. BST on August 17, 2026. Workshops are planned for November 2026, and another application cycle is expected to begin that month. (gov.uk)

Sources

• EMA, final reflection paper on real-world data in non-interventional studies, EMA/99865/2025. (ema.europa.eu)
• EMA, concept paper on external controls, EMA/125200/2026. (ema.europa.eu)
• MHRA, draft guideline and closed consultation on real-world-data external control arms. (assets.publishing.service.gov.uk)
• MHRA-NICE Real-World Evidence Scientific Dialogue and expression-of-interest guidance. (gov.uk)