Real-world evidence methods for regulatory decision making: An HMA/EMA update

Highlights and insights from the joint Heads of Medicines Agencies (HMA) / European Medicines Agency (EMA) Big Data Steering Group Workshop on RWE methods: Harnessing Real-World Data for Regulatory Use  

Regulatory-blog-image_100x100.jpgThis blog is part of The Regulatory Navigator series, where we explore the evolving regulatory landscape with actionable insight from Parexel's experts, sharing their experience to maximize success for clinical development and patient access.


To gather perspectives on the draft real-world evidence (RWE) reflection paper (RP)1, discuss priorities for future regulatory guidance development and collaboration and engage stakeholders regarding novel RWE methods in regulatory decision-making, the joint HMA/EMA Big Data Steering Group hosted a workshop. Held on the 14th of June at the EMA in Amsterdam, Parexel was invited to participate. Here, we share the key highlights and insights from the session.  

 1. Presentation and discussion of RWE RP  

To enhance the reliability of the evidence, the paper identifies potential limitations and proposes strategies to mitigate them by focusing on the design, conduct, and analysis of non-interventional studies (NIS) using RWD.  

 Key points in the paper include: 

  •  Study design: Should primarily be driven by the need to obtain reliable evidence for the research question at hand.  
    • Feasibility assessments for proposed study designs should be discussed with regulatory authorities. 
    • The Estimand framework2 should be considered when designing hypothetical target trials. 
  • Bias and confounding: 
    • For studies with a causal objective, it is crucial to address the risk of selection bias, information bias, and confounding. 
    • Adequately define and justify inclusion/exclusion criteria to minimize bias and confounding. 
    • If an external or historical cohort is used for comparison, the decision should be justified, and attention should be given to potential differences in selection criteria that may introduce bias and confounding. 
    • The pathway of data collection and extraction should be identified, and different steps must be verified to mimimise the risk of information bias. 
    • Time-related bias and confounding identification were also highlighted. Consideration should be given to the effect modifiers. 
  • Data governance and quality:  
    • Sources of RWD used in a study should be transparent, and the principles of the European Network of Centres for Pharmacoepidemiology and Pharmacovigilance (ENCEPP) code of conduct3 should be applied. 
    • Sponsors are encouraged to register their studies in the HMA-EMA catalog.4 
    • For statistical analysis, the focus should be on estimation, hypothesis testing, and the inclusion of time-dependent analyses when appropriate for the research question in cohort studies where events occur at different time points.  
    • It is recommended to discuss the Statistical Analysis Plan (SAP) with regulators. 

 From a multi-stakeholder discussion following the presentation of the paper, key themes emerged:  

  • An alignment with key points covered in the RP, including the importance of feasibility assessments, transparency, patient access, methodology (specifically as relates to patient access), bias, and data quality. 
  • However, there are some areas for clarification and development required: 
    • The purpose of the RP 
    • A greater emphasis on analytical bias, and how the same data can yield different results  
    • The definition of non-interventional studies (NIS) 
    • The applicability of relevant text to both primary and secondary research 


  • Xavier Kurz (ESEC RWE) 
  • Almath Spooner (Pharmaceutical industry – EFPIA) 
  • Helga Gardarsdottir (Academia)  
  • Bettina Ryll (Patient perspective) 
  • Holger Schunemann (HCP perspective) 
  • Olaf Klungel (moderator) 
2. RWE methods to support EU regulatory decision-making: RWD-derived external controls in clinical trials 

Through multi-perspective discussions and case studies, the challenges, considerations, and potential benefits of using RWD-derived external controls were presented.  

  • Considering the use of RWD, NIS and randomized clinical trials (RCT) as data sources for external controls in regulatory decision-making: 
    • RCTs provide comparative multiple endpoints, but there are also challenges in their interpretation due to uncertainties.  
    • RWD as a control adds more uncertainty, complicating the assessment.  
    • Alternative designs with less uncertainty are often scientifically justified and preferable. 
  • RCTs are preferred whenever possible, external control data can be acceptable in situations like high unmet medical needs or rare diseases (FDA perspective through presentation of a use case). 
  • There is a focus on recognizing and understanding limitations regarding the use of external controls in RWE studies and understanding the contextual use of external controls to contextualize the results of pivotal studies (EMA perspective through the presentation of a use case) 
    • Limitations include: challenges related to unknown confounders, limitations of propensity scoring, assessing time-to-event endpoints in non-randomized comparisons, and the discouraged usage of p-values in the Summary of Product Characteristics (SmPC).  
  • Additional considerations for RWD-derived controls in clinical trials: 
    • Project Pragmatica5 at US FDA is a good example of integrating the best aspects of RCT and RWD 
    • Distinguish between retrospectively and prospectively collected high-quality data 
    • Validate external control arms through additional registries or similar methods due to uncertainties and biases  


  • Elina Asikanius (SAWP, MWP) 
  • Maurille Feudjo Tepie (UCB, industry perspective) 
  • Donna Rivera (FDA perspective) 
  • Andrea Buzzi (EMA)  
  • Theodor Framke (Methodology ESEC) 
  • Mehmet Burcu (Industry perspective) 
  • Denis Lacombe (Academic perspective) 
  • Jan Cornelissen 
3. Focus of the Methodology Working Party (MWP) 

The MWP is part of EMA’s Committee for Medicinal Products for Human Use (CHMP) working parties and is committed to maintaining high standards of evidence. Their future focus areas include external controls, Bayesian statistics, platform trials, the development of a guideline on predictive biomarker co-development, as well as the application of artificial intelligence (AI) in both clinical development and pharmacovigilance. Their roadmap for the development of RWE guidance includes providing guidance on external controls derived from RWD and using external data to supplement control arms in clinical trials. A panel discussion further highlighted the importance of involving patients and considering their perspectives throughout the drug development process, underscoring the significance of patient-centric approaches. 


  • Kit Roes (MWP Co-chair) 
  • Jeppe Larsen (BDSG Co-chair) 
  • Olaf Klungel (MWP member) 
  • Marieke Schoonen (Industry perspective) 
  • Viviana Giannuzzi (Academic perspective) 
  • George Paliouras (Patient perspective) 
  • Ioana Agache (HCP perspective)  
Next steps  

Impactful evidence can be generated through RWD, but it is critical to consider a spectrum that includes both RWD and RCT – as well as focusing on factors such as Treatment Policy Effectiveness (TFE) and estimands. An inclusive approach contributes to accelerating access to innovative treatments. 

The new RP is an important step in guiding industry and regulators, and Parexel will be providing feedback that aligns with our position that RWD and RWE can be used for regulatory decision-making, including in the pre-licensing phase, but it requires careful planning, early interaction and the necessary expertise. 

Parexel’s team of ex-regulators and industry experts are ready to discuss the impact of the RP and best practices outlined during this workshop on your RWD/RWE regulatory strategy in Europe. Please get in touch, we’re always available for a conversation.  



  1. Reflection paper on use of real-world data in non-interventional studies to generate real-world evidence - Scientific guideline: 

  1. Estimands—A Basic Element for Clinical Trials:,and%20handling%20of%20intercurrent%20events

  2. ENCePP Code of Conduct: 

  3. HMA-EMA Catalogues of real-world data sources and studies: 

  1. Project Pragmatica: 


Return to Insights Center

Related Insights