Real-world evidence

Real-world evidence comes from data generated in ordinary care — electronic records, registries, claims, pharmacovigilance reporting — rather than from a trial's controlled conditions. It can answer questions trials are poorly suited to: long-term safety, rare harms, and how a treatment behaves in the patients who were never eligible for the study.

It cannot substitute for randomisation when the question is whether a treatment works. Our reporting keeps that boundary visible, because regulators are increasingly willing to accept such evidence and the standards for it are still forming.