(SLC-043) Utilization of Real-World Evidence in Drug Approvals at the European Medicines Agency from 2020 to 2023

Background: Integrating real-world evidence (RWE) into regulatory science is vital for addressing the evidence gaps left by randomized controlled trials (RCTs), which do not capture the complexities of routine clinical practice. However, the extent to which RWE influences the approval process at the European Medicines Agency (EMA) remains inadequately understood.

Objectives: To determine the extent to which RWE was integrated into EMA regulatory assessments for new medicines between 2020 and 2023, and to assess its impact on approval outcomes across therapeutic domains.

Methods: A systematic three-step research process was used. Step 1 identified EMA assessment reports based on predefined criteria. Step 2 involved targeted keyword screening to identify relevant RWE studies. Step 3 included systematic data extraction of 17 variables per study. Two researchers independently reviewed the data, with final validation by a senior researcher.

Results: Of the 423 identified reports, nearly all included planned RWE studies; however, only 39 (9%) incorporated RWE in the approval process. These 39 reports included a total of 68 relevant RWE studies. 20 studies were excluded from further data extraction because their reports did not provide sufficient information.Over time, the number of RWE studies varied, with 20 (18%) in 2020, 24 (21%) in 2021, 13 (12%) in 2022, and 11 (13%) in 2023. Oncology (38%) was by far the leading therapeutic area, followed at a distance by COVID-19 (15%) and neurology (13%), though other fields were represented as well. Despite a notable increase in RWE submissions during the COVID-19 pandemic—likely driven by challenges in conducting traditional RCTs and the need for timely evidence—63% of withdrawn studies were in the therapeutic area of COVID-19. Most of these studies employed a cohort design, with objectives spanning descriptive, safety and efficacy/effectiveness aims. Data sources included registries (22; 46%), electronic health records (16; 33%), claims databases (3; 6%), pharmacovigilance databases (1; 2%), and (11; 23%) unspecified sources. Explicit regulatory endorsement remained limited, appearing in only 29% of reports that referenced RWE.

Conclusions: While nearly all assessment reports included planned RWE studies, only 9% were used in the approval processes. Although RWE can help capture the complexity of clinical practice, it is essential that this evidence is presented in a structured and standardized manner, similar to the approach used for RCTs in the reports. Clearer and more standardized EMA guidelines will help to enhance further RWE in regulatory use.