Platform endpoints, estimands and comparisons

Define arm-specific questions and populations

Each comparison needs its own estimand, eligible population and claim family. A shared platform does not imply that every arm is comparable with every participant in the control pool.

Flow diagram

Platform endpoints, estimands and comparisonsF1.S1Define master and arm objectivesF1.S2Define each comparison populationF1.S3Specify estimands and endpointsF1.S4Define claim families andmultiplicityF1.S5Version comparison specificationsF1.N1Platform comparison registerComparisonEligiblepopulationEndpointEstimandEpochClaim familyPrimarycomparisonSecondarycomparisonSafetyF1.N2Assumptions and rationaleF1.N3Method and protocol evidenceF1.S0 Start flowPlatform endpoints, estimands andcomparisonsF1.S999 End flowEnd flow

Steps

Define master and arm objectives

Separate common endpoints from arm-specific scientific questions.

Define each comparison population

Specify cohort, eligibility, epoch and comparator pool.

Specify estimands and endpoints

Record outcome timing, intercurrent events and summary measures.

Define claim families and multiplicity

Justify error-control scope for related versus distinct claims.

Version comparison specifications

Preserve protocol amendments and avoid retrospective changes to primary questions.

Platform endpoints, estimands and comparisons

End flow

Tables and supporting information

Platform comparison register

Platform comparison registerComparisonEligiblepopulationEndpointEstimandEpochClaim familyPrimarycomparisonSecondarycomparisonSafety

Each comparison needs its own target and control eligibility definition.

Assumptions and rationale

There is no universal multiplicity correction for every platform; the claims and comparison structure determine the strategy.

Method and protocol evidence

Attach the protocol, methodological sources and versioned software specification supporting these decisions. Record citations and provenance; do not enter participant identifiers.

Sources