UK’s MHRA issues draft guidance on use of RWD external control arms in clinical trials

A new draft guideline from the UK Medicines and Healthcare products Regulatory Agency offers guidance for sponsors on designing clinical trials that use external control arms built from real-world data, with the aim of supporting regulatory submissions.

Coinciding with International Clinical Trials Day (May 20, 2025), the UK Medicines and Healthcare products Regulatory Agency (MHRA) has launched a public consultation on its draft guideline for the use of real-world data (RWD) external control arms (ECAs) in clinical trials. This initiative is part of a broader effort to innovate trial methodologies and enable more flexible, efficient approaches to evidence generation, which are particularly relevant in contexts where randomized controlled trials (RCTs) may not be ethical, feasible or timely.

The 6-week consultation invites stakeholder feedback on the clarity and content of the draft guideline, which outlines key considerations for incorporating RWD-derived ECAs into trial designs intended to support regulatory decisions. While the guideline provides general principles, it is not exhaustive. Sponsors seeking more detailed input are encouraged to request a scientific advice meeting, which can include MHRA representatives with expertise in licensing, clinical trial approval, post-licensing studies, pediatrics, medical devices, inspections, and electronic healthcare records (EHRs). As the guidance notes: “Sponsors interested in the use of RWD in their development programs are encouraged to engage with the MHRA for further advice on specific proposals.”

RWD is defined in the draft as, “data relating to patient health status or delivery of healthcare collected outside of a clinical study.” This includes EHRs, disease registries, administrative datasets, patient-reported outcomes (PROs), and data from wearable devices.

The guidance outlines two primary use cases: single-arm trials that rely solely on an ECA, and underpowered RCTs that are augmented with an ECA to improve statistical power or contextualize findings. A small internal control arm supported by RWD may provide greater confidence than relying solely on external data. “Consistency with the randomized internal control arm in terms of results provides some reassurance that the external dataset has been chosen appropriately,” the guidance notes.

While fully powered RCTs remain the preferred design, “there is no general scenario where the use of RWD external controls is explicitly ruled out.” Instead, RWD ECAs may be acceptable where an RCT is not viable – such as in rare diseases or conditions with no existing standard of care. In such cases, trial protocols must meet the same standards as traditional RCTs, including pre-specifying objectives, endpoints, and analysis methods.

The MHRA emphasizes the need for methodological rigor: “Given that there is no randomization or blinding, it will be necessary to address issues relating to bias when using an ECA.” They recommend several strategies to mitigate bias include close alignment between ECA and trial populations, robust justification of data sources, and clear definition of index dates and follow-up periods. Use of concurrent data is preferred to historical data to minimize time-related biases such as immortal time bias. Where retrospective inclusion of an ECA is necessary, sponsors must demonstrate that the choice was not result-driven.

The guideline also highlights the importance of aligning estimands between the trial and the ECA, referencing the estimand framework, as described in ICH E9(R1). In addition, sponsors are advised to consider the extent and handling of missing data and to conduct sensitivity analyses to explore the robustness of their findings.

This draft, which complements the MHRA’s existing guidance on RWD use in clinical trials, is open for consultation until June 30, 2025. Stakeholder responses are intended to refine the final guidance, supporting the MHRA’s goal of enabling innovative yet scientifically sound trial designs that enhance regulatory confidence and improve patient relevance.