EMA adopts reflection paper on RWE from noninterventional studies
The European Medicines Agency (EMA) has adopted a reflection paper on the design and conduct of non-interventional studies (NIS) that generate real-world data (RWD) for regulatory purposes.These studies can be used to characterize a disease's epidemiology, describe care standards, conduct post-marketing monitoring, and investigate safety issues.
The paper discusses the legal obligations and regulatory requirements governing NIS studies, the design of NIS studies, bias and confounding, governance and transparency, data quality, and statistical analysis.
The paper was adopted on 17 March, and it contains minor modifications from the draft issued in May 2024. (RELATED: Euro Roundup: EMA shares draft paper on non-interventional studies using RWD, Regulatory Focus 9 May 2024).
The update includes a new chapter on bias and confounding, along with a new section on model specifications in the statistical analysis section.
In the reflection paper, the EMA defines NIS as a clinical study that does not fulfill the criteria that characterize a clinical trial. Clinical trials typically employ randomization, blinding, and controlled environments as their primary methods for assessing a drug's safety and efficacy. In contrast, NIS use RWD gathered from routine medical practice, where patients are treated with marketed drugs and are not assigned to specific interventions.
The guidance states that “the ability to capture electronic healthcare data and data from registries is now providing new opportunities to use RWD in NIS and generate RWE that reflects clinical practice. As such, NIS using RWD can complement and support data from RCTs by filling gaps in knowledge and reducing uncertainties about a product’s safety and effectiveness.”
EMA emphasized that quality of the data is a key factor in assessing the appropriateness of RWD for regulatory use. This includes the reliability and relevance of the data, as well as how accurately the data represents standard clinical practice based on the research question.
EMA said that “reliability determines whether data represent the intended underlying medical concepts and are complete, trustworthy, and credible. Different dimensions of reliability can be evaluated and documented by using a data quality framework.”
The agency recommends using the framework spelled out in the Heads of Medicines Agencies (HMA)-EMA Data Quality Frameworks for EU medicines. This framework should be adapted to fit the specific data source, and it is important to develop expertise to implement at least one data quality framework in the guidance.
EMA defines relevance as follows: it refers to whether key data elements—such as exposure, outcomes, and covariates—are available for a specific research question. Additionally, it considers whether the size of the study population is adequate, whether the population is representative and sufficiently covers the target population for the study objective, and whether the study design is appropriate to effectively answer the research question.
The agency recommends use of the HARmonised Protocol Template to present information on the study design, exposure, outcome, and covariates and to evaluate RWE studies.
The update adds a new chapter on bias, confounding, and effect modification. New text states that “the non-interventional nature of NIS may lead to bias that distorts the measure of association due to processes of selection….and confounding (difference in the risk of developing the outcome of interest). If not adequately prevented or controlled, bias and confounding may limit the causal interpretation of the results. Potential effect modification of interest needs to be addressed to identify and describe study population characteristics that may affect the generalisability of the study results to a defined target population.”
It also has a new section on model specifications under the statistical analysis section. It states that “the assumptions of the analytic approach (specification of the statistical model, key variables, censoring assumptions, etc.) and the model diagnostics are important considerations to be presented in the statistical analysis plan. These assumptions, and their possible impact on the results, should be addressed through sensitivity analyses and assessed in the final study report.”
EMA has also published a roadmap detailing its progress towards establishing regulatory guidance for RWD. Efforts to incorporate RWD into the medicine approval process in the EU began in 2022. Since then, the annual work plans of the Committee for Medicinal Products for Human Use (CHMP) have included activities related to real-world evidence (RWE). This published work plan addresses a recommendation made in the CHMP's 2022-2024 work plan, as well as the current 2025-2027 work plan of the Methodological Working Party (MWP).
EMA reflection paper
×
Welcome to the new RAPS Digital Experience
We have completed our migration to a new platform and are pleased to introduce the updated site.
What to expect: If you have an existing login, please RESET YOUR PASSWORD before signing in. After you log in for the first time, you will be prompted to confirm your profile preferences, which will be used to personalize content.
We encourage you to explore the new website and visit your updated My RAPS page. If you need assistance, please review our FAQ page.
We welcome your feedback. Please let us know how we can continue to improve your experience.