Product review ledger

How the product manager should review

For each row, choose Accept, Change, Split, Merge, or Need scenario. A change is not complete until the reviewer supplies a counterexample or revised rule.

ID	Assertion under review	Decision	Counterexample / proposed wording
R-001	Study identity survives protocol amendments and identifier corrections	Unreviewed
R-002	Study Country owns country-specific participation and approval state	Unreviewed
R-003	Study Site is Study × Facility participation	Unreviewed
R-004	Site activation and permission to enroll are distinct	Unreviewed
R-005	Participant identity is study-scoped by default	Unreviewed
R-006	Consent withdrawal and study withdrawal are independent decisions	Unreviewed
R-007	Enrollment requires consent and eligibility under one governing version	Unreviewed
R-008	Planned encounter and actual Subject Visit never share identity	Unreviewed
R-009	CTMS consumes subject milestones rather than owning EDC clinical values	Unreviewed
R-010	A data review decision applies to an exact data revision	Unreviewed
R-011	Query closure and data correction are separate behaviours	Unreviewed
R-012	Protocol deviation requires a classification decision against a rule	Unreviewed
R-013	Monitoring Visit and Trip Report have separate lifecycles	Unreviewed
R-014	Monitoring Finding may be promoted into, but is not automatically, a Quality Issue	Unreviewed
R-015	Essential Record and Document Version are distinct	Unreviewed
R-016	TMF filing scope is study, country, or site per record instance	Unreviewed
R-017	Participant visit/procedure occurrences can trigger payable items without Payments owning them	Unreviewed
R-018	Study responsibility never directly grants system authorization	Unreviewed

Required counterexample format

Given:      the exact starting objects and versions
When:       one named actor attempts one action
Then:       the outcome the model should permit
Because:    the clinical, regulatory, or operational rule
Evidence:   standard, SOP pattern, or known product behaviour

Example:

Given:   an active site under protocol v3 and a participant consented under v2
When:    the investigator enrolls the participant after v3 became effective at the site
Then:    enrollment is refused until the required v3 consent is effective
Because: enrollment evidence and the governing site version must agree

Exit criteria for v0.2

• Every spine object has an accepted identity statement.
• Every cross-context fact has one accepted publisher.
• At least one happy path and two refusal scenarios pass for each aggregate.
• No unresolved disagreement is hidden in a generic status, JSON attribute, or “configurable” field.
• Vendor terminology differences are mapped explicitly without inheriting vendor storage design.