Table 2.

Four main statistical approaches to validate a surrogate end point

ConceptsApproachParadigmDataTreatment effectQuestion addressed
Proportion of treatment effect Single-trial Causal effect IPD Observed What is the proportion of treatment effect on OS, that is, explained by the surrogate? 
Trial-level coefficient of determination Multiple trial Causal association IPD Observed Based on past trials, how will treatment effect measured by surrogate predict the treatment effect on OS in a future trial? 
Principal stratification Single trial, or multiple trial Causal association IPD Potential (using counterfactuals) Granted the status of the surrogate, can the treatment assignment predict OS? 
Mediation analysis Single trial Causal effect IPD Potential (using counterfactuals) How is the treatment effect on OS mediated by the surrogate end point? 
ConceptsApproachParadigmDataTreatment effectQuestion addressed
Proportion of treatment effect Single-trial Causal effect IPD Observed What is the proportion of treatment effect on OS, that is, explained by the surrogate? 
Trial-level coefficient of determination Multiple trial Causal association IPD Observed Based on past trials, how will treatment effect measured by surrogate predict the treatment effect on OS in a future trial? 
Principal stratification Single trial, or multiple trial Causal association IPD Potential (using counterfactuals) Granted the status of the surrogate, can the treatment assignment predict OS? 
Mediation analysis Single trial Causal effect IPD Potential (using counterfactuals) How is the treatment effect on OS mediated by the surrogate end point? 
Close Modal

or Create an Account

Close Modal
Close Modal