Study model · Protocol v7 · Amendment 02
One study, one meaning.Every part still itself.
Capture, randomization, coding, signature and export stay distinct, and read one study model. Resolve an amendment once; every part reads the new version.
- Formsv7
- Edit checksv7
- Visit schedulev7
- Randomizationv7
- Exportv7
Three layers
A trial makes three kinds of thing.They have to agree.
Read once, not re-typed.
Protocol, amendments, specifications and SOPs become one model. Every concept keeps a pointer to its line.
Protocol v7 · §6.3 · p.52Visit 3 occurs on Day 28 ± 5 days after randomization.
In the modelVisit 3 · window ±5 days→ §6.3 p.52
Every choice, attributable.
Eligibility, a deviation, a query, a signature: each timestamped and tied to the clause that authorized it.
Generated, not configured.
No mapping step between the model and the parts that run it.
- CaptureeCRF builder · protocol parser · validation
- OversightQueries · SDV · double entry
- AllocationRandomization · blinding · supply
- TerminologyMedDRA · WHODrug
- ClosureLocks · e-signature · export
One platform
Every seam is a mapping.Every mapping can disagree.
4 mappings · four places to disagree about one subject
0 mappings · one meaning, read by every part
A CTMS or a dedicated safety database is integrated over a REST API, not replaced.
The loop
Build, monitor, amend, prove.Then again.
01 · Build
The protocol drafts the build.
Forms, checks and the visit schedule. A person reviews and applies; that act puts it live.
8visits6forms54fields70rules16branches02 · Monitor
Entry is read against the protocol.
Not generic bounds. A missed assessment surfaces while it can still be fixed.
Visit 3Recorded Day 33 · window Days 25–31 deviationRaisedAgainst §6.3 · protocol v4.0 cited03 · Amend
An amendment states its consequences.
Which forms change, which windows move, which findings are affected.
~Visit 3 window ±3 → ±5 days2forms4checks128values04 · Prove
The evidence is a replay.
Each movement wrote its reasons as it happened. Nothing is reconstructed at the end.
10:02Signed · bound to version 3 investigatorExportODM XML with the audit trail readyWhile it runs
Arms change mid-study.Nothing is rebuilt.
Add, stop, graduate or reopen an arm as a reviewed, signed change. Collected data stays where it is, and reconciliation runs as you go.
FAQ
Clinical trial software,answered plainly.
What is the difference between EDC and clinical trial software?
EDC software is the data layer: form design, entry, validation and export. Clinical trial software is the wider platform around it, adding randomization, query management, medical coding, source data verification, e-signature and study administration. Both are one product here.
Can this replace several clinical trial vendors?
In most study designs, yes: capture, randomization and supply, queries, coding, SDV, signature, locks and regulatory export are all in scope. For a CTMS or a dedicated safety database, the answer is integration over a REST API rather than replacement — we do not build those and we will not pretend otherwise.
How are protocol amendments handled?
As a versioned change to the study model. Update the forms, fields and affected rules, review the stated impact, then activate the new version. Because every part reads the same model, the change propagates without a mapping exercise between vendors.
Can arms be added, stopped or changed while the trial is running?
Yes, without rebuilding the study. A new or changed arm arrives as a reviewed amendment, with every form, check and collected value it touches counted before anyone signs. Stopping, graduating or reopening an arm is proposed, reviewed by a different qualified person and executed with an electronic signature, and participants already on the arm keep their assignment, follow-up and data.
How is trial data reconciled?
Continuously, inside the study. Adverse events are reconciled against safety reporting case by case, with each discrepancy resolved and verified by a second investigator; drug-kit counts are reconciled site by site with signed evidence; and every import is reconciled against its source. Open safety items hold closeout until they are resolved.
Bring the stackyou are trying to replace.
Thirty minutes on the real platform, against your study type rather than a demo study.
Or email us: jgui@accuratrials.com