EMA details data quality considerations when using RWD

The European Medicines Agency (EMA) has finalized a document with recommendations on using the European Medicines Regulatory Network (EMRN) Data Quality Framework (DQF) when submitting premarket applications that include real-world data.

The agency said the recommendations in the document are intended to help regulators evaluate the quality of data used to assess real-world evidence supporting marketing applications.

"The EMRN DQF sets out the principles, concepts, and definitions as intended to be applied widely across datasets used in medicine regulatory use cases," said EMA. "It also provides examples and in-depth clarifications on the developed framework elements for characterising, assessing, and assuring Data Quality (DQ) in the regulatory context.

"It is therefore recommended to use the EMRN DQF as a companion document when reading this document," the agency added. "Unless otherwise specified, the definitions provided in the EMRN DQF also apply to this document."

EMA explained that the scope of the document covers RWD collected during clinical practice to answer specific research questions and follows the European Network of Centres for Pharmacoepidemiology and Pharmacovigilance (ENCePP) Methodological Standards in Pharmacoepidemiology and the ICH M14 Guideline. Some examples of RWD that can be used in this context include administrative or claims data, electronic medical records (EMR), electronic health records (EHR)), pharmacy and prescription data, and patient registries.

The document does not cover RWD from repurposed previously published analyses, such as meta-analyses, quality of certain specialized data, such as direct-from-patient data, and comparative assessment of data quality based on data origin or collection approach.

EMA has detailed the responsibilities of data holders, data users, and regulators in a table in the document. The agency notes that while data holders are responsible for aspects such as implementing the quality management system (QMS) that governs the data, ensuring data quality metrics, and publishing the metadata, the data users are responsible for assessing whether the data is fit-for-use, documenting data quality metrics, and validating and interpreting the data. Regulators' responsibilities include providing guidance on data quality, reviewing the data validation process, and interpreting the data.

Another important aspect of the document is that it lays out EMA's thinking on data quality considerations for the secondary use of RWD. The agency notes that secondary use of RWD data needs to be fit-for-use, depending on the research questions, and comprehensive metadata documentation, data enrichment techniques, and statistical imputation can improve its usability. It also emphasized the importance of anonymizing and pseudo-anonymizing data, since it was not originally captured with the intent to use it in a marketing application and may sometimes require obtaining retrospective informed consent from people in the data sets.

"In a secondary use of data scenario, there is no possibility to control most DQ factors of reliability (e.g., accuracy and precision) at the source (point of data recording), and even the ability to assess reliability is limited," said EMA. "Therefore, the primary focus of DQF implementation is the identification of reliability gaps that could lead to record removal or amendment with approximated values (only in some limited cases it can lead to the correction of the data capture processes) and to maximise the amount of information supporting a reliability assessment."

Besides considering the reliability of the RWD when it is being considered for a secondary use, EMA also highlighted that applicants should consider the impact of the data's extensiveness and representativeness, as well as its coherence and timeliness. The agency also emphasized that it is important to ensure that data collected from RWD is properly collected and maintained so that it can be used reliably for regulatory purposes.

"The quality of RWD cannot be assured unless the systems and processes responsible for their collection or recording and transformation are reliable and offer the necessary guarantees," said EMA. "If RWD are considered for regulatory use, no matter the content of an RWD source, its use would be feasible if there is sufficient evidence that the information provided is true and not accidentally or intentionally altered."

EMA details what it expects from the systems and processes used to collect and manage RWD data that could give the agency that reliability. More specifically, it addresses maturity models and other considerations to ensure data quality and usability.

EMA document