External comparators and estimands

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Observational Research

Background:

• The ICH E9(R1) addendum provides an estimand framework to define treatment effects clearly.
• This framework specifies attributes like treatment conditions, population, endpoints, handling of intercurrent events, and population-level summaries.

Purpose of the Study:

• To explore the implications of hybrid clinical/observational research, specifically using External Comparators (ECs), on the existing estimand framework.
• To identify potential challenges and necessary modifications to the estimand attributes when applying the framework to ECs.

Main Methods:

• Conceptual analysis of the ICH E9(R1) estimand framework in the context of External Comparators.
• Identification of specific challenges related to treatment conditions, population definition, endpoint comparability, intercurrent events, and population-level summaries in EC studies.

Main Results:

• External Comparators present unique challenges for defining treatment conditions, populations (moving beyond classical Intention-to-Treat), and ensuring endpoint comparability.
• Handling of intercurrent events and the applicability of the hypothetical treatment policy become more critical, especially for long-term endpoints.
• Data heterogeneity in ECs can challenge assumptions for population-level summaries, such as proportional hazards.

Conclusions:

• The current estimand framework requires careful consideration and potential expansion to adequately address the complexities of External Comparators in hybrid research.
• Baseline definition and marginal estimators are proposed as potential additional estimand attributes for future revisions.
• Adapting the estimand framework is essential for robust and reliable treatment effect estimation in studies involving external data sources.