Biometrics partner for sponsors and CROs

Our process

Four stages from kick-off to the questions and reports that follow a submission, with every step documented and traceable.

1

Plan

We start by reading the protocol and the case report forms with the study team, so the data will support the analysis. The statistical analysis plan and table shells are agreed before any outputs are programmed, and the SAP is finalised before the blind is broken.

  • Kick-off: scope, timelines, standards and who signs off what.
  • Protocol and CRF review from a programming point of view.
  • Statistical analysis plan and mock table shells.
2

Build

Data from EDC systems, central labs, safety databases and external vendors is mapped to SDTM from agreed specifications, then turned into ADaM analysis datasets with derivations that trace back to the source. Dry runs on blinded data find problems while there is still time to fix them.

  • SDTM and ADaM specifications, controlled terminology and annotated CRFs.
  • SDTM and ADaM datasets, independently validated.
  • Dry runs of tables, listings and figures on blinded data.
3

Deliver

After database lock and unblinding, we run the final tables, listings and figures, support the statistical sections of the clinical study report, and build the integrated analyses and the e-submission package.

  • Final tables, listings and figures, double-programmed where required.
  • Integrated summaries of safety and efficacy (ISS and ISE).
  • e-Submission package: datasets, define.xml, cSDRG and ADRG.
4

After submission

Review brings questions, and approval brings regular safety reporting. We provide the programming support for both, reusing the validated datasets and programs from the submission so answers are fast and consistent.

  • Health authority questions: new analyses and outputs within the agency's deadline.
  • DSURs and PSURs/PBRERs: exposure summaries, tabulations and line listings.
  • ISS and ISE updates as new data or questions arrive.

Have a study that needs statistical support?

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