Biometrics partner for sponsors and CROs

Getting an e-submission package ready when validation found hundreds of issues

Weeks before a planned submission, conformance checks returned a long list of findings. We fixed what could be fixed and explained the rest in the reviewer's guides.

1

The challenge

The sponsor's datasets had been built by several vendors over the life of the programme. When the package was run through conformance validation shortly before submission, it produced a long list of errors and warnings, inconsistent define.xml content across studies, and no cSDRG or ADRG.

2

What we did

We triaged every finding into three groups: genuine data or mapping errors to fix, findings caused by how the data was collected that could not be fixed, and checks that did not apply. Errors were corrected through the specifications and the datasets rebuilt, define.xml was regenerated in version 2.1 with consistent metadata, and we wrote the cSDRG and ADRG, giving a clear explanation for each finding that remained.

3

The result

The sponsor submitted on its planned date with a complete package: corrected datasets, define.xml, and reviewer's guides that answered the obvious questions before a reviewer had to ask.

Have a study that needs statistical support?

Tell us about the study and what you need. A statistician will get back to you.