Brief summary from "Final advice to EMA"

For the transparency of clinical data, advisory group discussed on the proactive publication of clinical trial data in EMA (European Medicines Agency). Needless to say, data content and format are the key issues of publication. The advisory group consists from 5 subgroup. Group 2, called CTAG2, summarized final advice on clinical-trial-data formats. For details, please follow the link.

When we talk about data formats for clinical studies, CDISC standards would be important, substantial and remarkable. This text tries to clarify some points regarding CDISC standard in the EMA final advice.

What are the differences?:
We have to be careful that this advice is for EMA, not for FDA. The biggest difference between EMA and FDA is the use case of clinical data. Efficacy for approval is the goal of FDA, sharing the data is the goal for EMA on the other hand. Thus, practical approaches are recommended for EMA. The uniqueness is that the preparation workload should not be burden for data provider. Not only big pharmas, but also small to middle sized pharmas and academia will be subjected to the requirements. The point is to collect clinical data for public. Easy-to-follow rules are important to realize the maximum data collection.

Data contents to be public:
One of our concern is data content to be open to the public. Table 1 in the section 3 illustrates what kind of data will be public in the program. Notably, "annotated Case Report Forms", "Patient-level dataset" and "Dataset specifications" are listed. All items are categorized as level 3. They are not required by EMA. No format is specified.

Any format is acceptable?:
In the final advice document, it's mentioned that data format consistency is not mandatory, though unified format is preferable. The importance of meta-data is referred. In this context, CDISC standards (especially SDTM, ADaM and define.XML) are suitable to handle these information. As FDA will mandate the use of CDISC standards near future, mega-pharama will create SDTM dataset for FDA submission. The datasets can be reused for this EMA requests. This is less resource consuming and consistent with attitude of the subgroup. However, CDISC standards may be burden for small-sized organization. Simple scanned image of CRF may be suitable for small clinical studies.

Our opinion:
CDISC standards are important as FDA submission component. As they are more common in the industry, other stakeholders consider that CDISC is de facto and well agreed standards. This adds new value for CDISC standards and it's getting more important for all data managers to understand the CDISC.





Author:TKD + SMZ