Statistical programming and data management for clinical sponsors
Under an FSP model you buy a function, not a whole study from a CRO. Our people join your biostatistics and data management teams, use your macro library and your EDC, and every deliverable goes through your QC before anyone signs it.
| Model | functional service (FSP) |
|---|---|
| Functions | programming, data management |
| Standards | CDISC SDTM, ADaM, Define-XML |
| Works in | your SAS or R, EDC and SOPs |
| Terms | written team proposal |
- 01Call
- 02Team proposal
- 03Onboarding
- 04Delivery
Scope
What the team takes on
Statistical programming
- SDTM. Raw and EDC data mapped to SDTM domains to your specifications, with the annotated CRF.
- ADaM. Analysis datasets derived from the SAP and dataset specifications.
- Tables, listings, figures. TLFs to the mock shells, for the CSR, safety reviews and DSMB packages.
- Submission package. Define-XML, reviewer's guides and validator findings closed or explained.
- Independent QC. Double programming and comparison on the outputs your QC plan names.
Clinical data management
- Study setup. Data management plan, eCRF and EDC build specifications, edit check specifications.
- Data cleaning. Data review, query management and reconciliation of external data such as labs.
- Coding. Adverse events and medications coded with MedDRA and WHODrug.
- Safety reconciliation. Serious adverse events reconciled between the clinical and the safety database.
- Database lock. Lock checklist, final reconciliations and the data transfer to programming.
Fit
When FSP support fits
Use it when you
- run studies in house and your programming or data management team is short before a database lock or a submission
- want one stable team across a portfolio of studies, not a new vendor for each study
- need people who work to CDISC and to your own macro library and SOPs from the first week
- want the routine units to move faster under the same QC, with an AI harness your programmers control
Skip it when
You want a CRO to run the whole study, including sites, monitoring and medical writing. We take on programming and data management only.
With AI
The same team, with a harness when you want one
Clinical programming already has its own checks: specifications, validator reports, log rules and independent double programming. Our AI harness for clinical programming cuts the work into units, such as one SDTM domain, one ADaM dataset or one table, runs each unit through those checks and keeps a run record. A named programmer from your team or ours signs every unit before it is released.
FAQ
Common questions
Q01What is an FSP model?
A functional service provider takes on one function of your clinical development, such as statistical programming or data management, across your studies. You keep the study, the standards and the sign-off. A full-service CRO runs the whole study instead.
Q02Do you work to our SOPs or to yours?
To yours. Our people are trained on your SOPs, standards and macro library before they touch a study, and the training records go into your files.
Q03SAS or R?
Either. We work in your licensed SAS environment or in R with pharmaverse packages, whichever your study and your submission plan use.
Q04Who owns the programs and datasets?
You do. Programs, datasets, outputs and documents live in your systems from the first day, and nothing is kept on our side.
Q05Can the team use the AI harness?
Yes, if you want it. The same team can run units through the clinical programming AI harness, with every output checked and signed by a named programmer.
Related
Related pages
- AI harness for clinical programming
Unit-based clinical programming: frozen inputs, pinned CDISC standards, automatic checks and a named signer per unit.
Service - Engineers for drug-trial statistics
Engineers with a healthcare statistics stack on a US medical technology project that runs statistical analysis for drug trials.
Case - Engineers in your team
Senior engineers who join your roadmap, in your tools and your process.
Direction
Next step
Tell us about your studies
A 30 minute call with Oleg about your studies, your standards and where the team is short. After it you get a written team proposal.
Oleg Pun
CEO, Vareger
