Quick answer: ICH E6(R3) Appendix C shifts the requirement from "essential documents" to "essential records" meaning sponsors must now identify, govern, and retrieve the documents, metadata, and data that let regulators reconstruct exactly what happened in a trial, wherever those records actually live (CTIS, CTMS, eTMF, safety systems, QA repositories, provider platforms). A static, single-system TMF checklist can tell you a file exists. It can't tell you whether the essential evidence chain behind a decision is intact and traceable across systems. Cloudbyz closes that gap with a native CTMS↔eTMF connection on Salesforce, run by an AI eTMF Agent that keeps essential records classified, tagged, and inspection-ready continuously not once a year.
Appendix C doesn't just add new paperwork. It changes what "control" means.
The old model: build a TMF Reference Model checklist, map artifacts to zones, chase completeness percentages, sample for QC. Green/amber/red dashboards stand in for actual control.
Appendix C asks a harder question: can you reconstruct what happened in this trial, using records that may not even be inside your TMF?
EMA's GCP Inspectors Working Group has flagged the same TMF failure modes for years missing essential documents, late filing, metadata errors, version control gaps, weak audit trails (2023 annual report, EMA PDF). A checklist confirms a document exists. It doesn't confirm the record is version-controlled, attributable, or retrievable in the timeframe a regulator expects.
Three things Appendix C now requires, that a checklist structurally cannot deliver:
Layer EU CTR and CTIS on top (EMA overview) and your essential records now span CTIS, CTMS, eTMF, safety systems, QA repositories, and CRO platforms simultaneously. A single-application checklist can answer "is Zone 03 complete?" It cannot answer "can we trace the essential records behind this decision, across every system that touched it?"
That's why most E6(R3) readiness work has quietly turned into a mapping exercise teams tracing which records sit where and how fast they can be produced. Mapping isn't governance. It's a symptom of not having a fabric.
Cloudbyz is built as a fabric, not a folder tree with a status column. The native CTMS↔eTMF connection on Salesforce means studies, countries, sites, and operational milestones share one governed data model and audit trail with TMF artifacts and workflows.
On that spine, the AI eTMF Agent operates continuously:
| ALCOA+ principle | How the CTMS↔eTMF fabric delivers it |
|---|---|
| Attributable / Contemporaneous | Human and AI actions on records and metadata sit in one Salesforce-native audit spine |
| Original / Accurate | The TMF points back to the authoritative source system (CTIS, safety, QA) instead of flattening it |
| Complete / Consistent | CTMS, eTMF, and linked repositories are reconciled continuously, not just before an inspection |
| Enduring / Available | Records stay locatable even as underlying systems and vendors change |
For EU portfolios, this means demonstrating E6(R3) compliance (effective 23 July 2025, EMA guideline) directly from your CTMS↔eTMF environment not a spreadsheet built the week before an inspection. For US portfolios, the same Salesforce-native foundation supports 21 CFR Part 11 expectations for electronic records and signatures.
When Appendix C comes up in an inspection room, "here's our document checklist" is no longer a sufficient answer. "Here's how our CTMS↔eTMF fabric keeps essential records navigable and inspection-ready as a matter of routine" is.
See what a CTMS-linked, AI-assisted eTMF looks like on your own portfolio.