Real-world evidence (RWE)

Real-world evidence (RWE) is clinical evidence about how a treatment, device, or intervention performs, produced by analyzing real-world data (RWD) — sources such as electronic health record systems, insurance claims, disease registries, and patient-reported outcomes — rather than data collected under the controlled conditions of a trial.

RWE differs from evidence generated by a randomized controlled trial because patients are not randomly assigned to treatment and control groups. Analysts instead rely on observational methods, matching, or other causal-inference techniques to estimate effects while adjusting for confounders such as disease severity or prior treatment history. Because assignment isn't randomized, RWE is more exposed to selection bias and confounding than a well-run trial, and results are typically described as associational unless the study design specifically supports causal claims.

Regulators including the FDA and EMA increasingly accept RWE to support drug approvals, label expansions, and post-market safety monitoring, especially for rare diseases or populations underrepresented in trials. It also underpins health economics and outcomes research and pharmacovigilance, where long-term, real-world safety and effectiveness matter more than a short trial window. The main pitfall is treating an association found in RWD as proof that the treatment caused the outcome, without adjusting for why certain patients received that treatment in the first place.