Advances in technology and changes to the regulatory environment have resulted in a change in drug applications, from being comprised of a collection of randomized clinical trials (RCTs) to RCTs plus real-world data (RWD) and data from various non-interventional study designs.
Most New Drug Applications (NDAs) and Biologics License Applications (BLAs) that were approved in the past two years include at least one real world or non-interventional study. However, our current clinical data submission standards (CDISC) were developed over 20 years ago and optimized for RCTs. These standards have not been updated to accommodate alternative study designs and RWD. This has presented numerous challenges to both sponsors and regulatory reviewers.
Current Clinical Study Design Standards
Current guidance dictates that RWD and data from other non-interventional study designs must be submitted using the standards documented in the FDA Data Standards Catalog. To date, this means that RWD must be submitted using CDISC standards; thus, sponsors need to convert real-world datasets into CDISC format when submitting this information in support of a marketing application. In the short term, this requirement has resulted in cumbersome data transformations, non-standard variables and domains, and business and validation rules which primarily apply to data collected in RCTs.
What else can we do, besides transforming data for all clinical study designs to comply with CDISC standards?
The industry should first consider changing the conversation from ‘How we fit data from non-interventional study designs into CDISC?’ to ‘What should the paradigm be for submitting data, when considering the increasing diversity of study designs?’
Can One Standard Accommodate All Clinical Study Designs?
Some topics for consideration include: Can one standard accommodate all study designs, or should a hybrid approach be employed? CDISC has developed core data elements for representing data from RCTs needed for regulatory submission. Can industry continue to develop core data elements and terminologies for non-interventional studies needed for regulatory submission and review?
Can CDISC standards be modified to accommodate data from all study designs or is the Observational Medical Outcomes Partnership (OMOP) Common Data Model or Fast Healthcare Interoperability Resources (FHIR) a more viable future submission standard? Over time, will more and more data be collected in EHRs? If so, does it make sense to transform this data from FHIR to another standard?
Finally, can our industry and regulatory authorities move beyond their comfort zone? Right now, CDISC is familiar to both sponsors and FDA. Sponsors and FDA are much less familiar with other standards such as FHIR and OMOP. Though CDISC standards are currently required, this should not inhibit our industry from considering the best future state and evaluating other alternatives. Collectively, we should continue to monitor how each standard is evolving and how this affects its ability to meet future submission needs.
Learn how you can define, reuse and govern standards in our Clinical MDR.