With FDA now regulating for all new studies to use CDISC standards when submitting data, it's vital that our processes are up to date with the latest standards and fit for purpose.
An FDA presentation at the recent PhUSE Computational Science Symposium tried to shed some light on how we're doing as an industry. It analysed all eCTD submissions from January-February 2017 that used standardized data.
You might be surprised how the submissions fared...
I was shocked to see that a massive 45% of submissions failed the basic conformance checks. Specifically, they didn't follow one of these basic rules:
- Must have DM dataset and CDISC Define-XML
- Must have ADSL dataset and CDISC Define-XML
All new studies must use standardized data, so there could be problems coming down the line.
It's not just clinical data
Another presentation gave data from the SEND "Fit For Use" pilot for non-clinical data showing similarly worrying results. Here 15% of submissions using standardized data would have failed the technical conformance criteria. Even more surprisingly, 62% of submissions had errors in their Define-XML files.
Define-XML has been around for twelve years now; SDTM even longer.
How can we help?
Standardized data means using CDISC standards, including Define-XML, SDTM, SEND and ADaM. If you're struggling with these, we can help.
Formedix are CDISC experts. We've been contributing to CDISC standards development for the last fifteen years. If you need help on CDISC means, how to do it quickly, or how to fix problems, we have everything you need.
Here are some of the ways we can help you:
- Validation of Define-XML
- Converting XPT datasets to Define-XML
- Defining CDISC SDTM, ADaM or SEND datasets and mapping data to them, using Define-XML
- Automated conversion of data to your CDISC datasets
- CDISC training, including eLearning
- Advice on how to produce standardized submissions
If you would like to know more, download our simple guide to how we can help you use standardized data.