UPDATE: As of November 2021, EMA has now issued its final guideline on registry-based studies, which offers recommendations in line with those described in its draft guidelines. Read on for our take on the draft guideline issued in September 2020.
—
With real-world evidence (RWE) growing in influence across health care, regulatory and health technology assessment (HTA) bodies continue to issue recommendations for biopharma to articulate standards around RWE generation and use.
However, as discussed during the session “RWE guidance and frameworks: What elements are necessary to promote quality, transparency, and validity in RWE?” at Virtual ISPOR 2020, these recommendations are often fragmented and pose challenges for manufacturers as they navigate different stakeholders’ nuanced definitions of “high quality RWE.”
In our “RWE guidance watch” series, we alert readers to recent, influential RWE guidances issued by standard-setting organizations and break down what they mean for biopharma organizations as they develop plans for RWE generation. The series begins with a look at a recent draft guidance from the European Medicines Agency (EMA); read on to learn more.
EMA guideline on registry-based studies
On September 24, 2020, EMA released a draft guideline on registry-based studies* to provide manufacturers with “recommendations on key methodological aspects that are specific to the use of patient registries.”**
EMA developed the guideline through consultation with various EMA committees and input from public stakeholders. A comment period is open until December 31, 2020.
Below, we share a summary of the key takeaways from these new recommendations.
Commitment to principled database epidemiology
A key theme present in EMA’s recommendations is a commitment to principled database epidemiology. For example, EMA notes that while registries can allow for comparisons between untreated and treated patients, registry-based studies—like all non-randomized studies—are prone to bias. The agency goes further and recommends specific methodological solutions to limit various types of bias, such as incorporating a time-dependent definition of exposure to classify immortal person-time and employing a new user design to mitigate the effects of immortal time bias and selection bias, respectively. Both recommendations are cornerstones of principled database study design.
In addition to specific methodological strategies to limit bias, the EMA guidelines highlight the need for transparency and encourage manufacturers to pre-specify the study protocol, data collection, data management, and data analysis plans. EMA also suggests that data reporting should be sufficient to replicate the study; however, it stops short of defining sufficient data collection practices.
In line with previous requirements for post-authorization studies, EMA calls on manufacturers to pre-register registry-based study protocols for post-authorization studies—a recommendation that is well aligned with the ISPOR-ICPE RWE Transparency Initiative. However, it does not explicitly call for the pre-registration of pre-authorization registry-based studies.
Focus on data quality and validity
While EMA’s recommendations range from study design and planning through statistical analysis, they focus heavily on data quality and validity.
EMA recommends that manufacturers utilize the REQuEST tool developed by EUnetHTA to evaluate the suitability of the registry for the research question of interest. EMA also recommends that a manufacturer collaborate with registry holders to conduct feasibility analyses before initiating a registry-based study. Manufacturers should evaluate data availability and quality, potential biases, and data privacy when building a registry-based study protocol.
Recommendation for manufacturers to seek scientific advice
While EMA discusses recommendations on how to evaluate registry data quality, fitness for purpose, and specific study design approaches to limit bias, it highly recommends that manufacturers utilize EMA’s scientific advice pathway. The pathway allows manufacturers to consult the agency on study design and methodological considerations when determining the appropriateness of a registry-based study.
Specifically, when manufacturers are planning to “deviate from a traditional [randomized controlled trial] design,” consultation with EMA (and HTA bodies, if appropriate) is recommended. As RWE increases in prominence and industry learns when and where RWE should be used in decision-making, this collaboration between EMA and manufacturers is necessary, and builds on the “learn by doing” approach that other groups are embracing in ongoing RWE demonstration projects.
Impact on biopharma
This EMA guideline is another iterative step in the agency’s work toward comprehensive guidance on RWE generation and use. It is limited in focus, only addressing registry-based studies, but likely alludes to key components of future guidelines on RWE. EMA is slated to publish guidelines on data quality and representativeness in 2022.
Research and development, value and access, and safety groups should all consider leveraging registry-based studies to supplement evidence generation for European regulatory and reimbursement decision-making. These groups may learn from the use cases for registry-based studies defined in the guideline, such as illustrating standards of care, determining incidence rates and determinants of disease outcomes, describing the characteristics of target populations, and helping validate surrogate endpoints. Additional use cases include supplementing a medicine’s value proposition and monitoring long-term safety and effectiveness.
Ultimately, when implementing any registry-based study, manufacturers should plan for consultation and collaboration with EMA and ensure alignment with the core pillars of principled database epidemiology.
* EMA defines a registry-based study as “an investigation of a research question using the infrastructure of (a) new or existing registry(-ies) for patient recruitment and data collection.”
** EMA defines a patient registry is an “organised system that collects data and information on a group of people defined by a particular disease or condition, and that serves a predetermined scientific, clinical and/or public health (policy) purpose.”