Endpoints and analysis

The trial’s Plan Endpoints and Analysis tabs share one governed Statistical record.

• Trial Summary owns objectives, endpoint definitions, estimands and their clinical intercurrent events and strategies.
• Endpoints owns native computation records, typed scope, analysis roles, assumptions and event implementations. Each analysis editor shows its native method outputs; reporting display contracts remain separate.
• Plan Analysis solely authors shared populations, sample sizes, the multiplicity graph and explicit interim applicability, with affected analysis and endpoint references.
• Analysis Specifications remains the separate governed ADaM implementation owner for datasets, variables, parameters and source/result bindings.

Populations and cohorts

Define analysis populations once in Plan Analysis. Protocol cohorts remain in Trial Summary → Population; an analysis population may reference their stable identities. Native computations select existing population IDs, and the server validates their scope.

Trial Summary owns forward estimand references to endpoint, population, intervention and sensitivity analyses. Clinical intercurrent-event records are inline estimand content, rather than references to a retired Statistical event registry. Statistical renders the inverse relationships as context.

Review and delivery

Manual changes require the existing governed Save and justification. Documents receives a structured current-source projection; Exports produces pinned artifacts with findings. Neither an exported metadata package nor generated prose is automatically approved.

See Trial endpoints and analysis for authoring, shared decisions and native AC/DC interchange.