Endpoint hypothesis and method

Endpoint and estimand → objective → hypotheses or estimation target → method → assumptions

Flow diagram

Endpoint hypothesis and methodEndpoint hypothesis and methodF1.S1Endpoint and estimandprimary · unspecifiedF1.S2Objective and comparison scaleF1.S3Hypotheses or estimation targetF1.S4Method and analysis populationF1.S5Assumptions and diagnosticsF1.S6Missingness and sensitivity planF1.N1Analysis method specificationEndpointPopulationEstimatoror testEffectscaleSidednessMarginSoftwareEndpointF1.N2Hypothesis notationF1.N3Model specificationF1.N4Method evidence

Steps

Endpoint and estimand

Define the target population, comparison, outcome, window and summary measure.

Objective and comparison scale

Specify estimation, superiority, noninferiority or equivalence; define favorable direction and scale.

Hypotheses or estimation target

Specify null and alternative, test sidedness, margin if applicable or precision target.

Method and analysis population

Choose an estimator/test and population appropriate to design and endpoint.

Assumptions and diagnostics

Record independence, distribution, covariance, censoring or model assumptions as applicable.

Missingness and sensitivity plan

Align handling of missing outcomes and sensitivity analyses with the estimand.

Tables and supporting information

Analysis method specification

Analysis method specificationEndpointPopulationEstimatoror testEffectscaleSidednessMarginSoftwareEndpoint

Leave method-specific settings blank until specified.

Hypothesis notation

Enter study-specific null and alternative hypotheses or a precision target. Do not assume that superiority and noninferiority use the same null.

Model specification

Record the model and its assumptions for the selected outcome.

Method evidence

Cite the protocol and methodological source.

Sources