Investigating estimand considerations in adaptive trials: a systematic review

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Regulatory Science

Background:

• Randomised controlled trials (RCTs) are crucial for evaluating treatments, but ambiguous research questions hinder interpretation.
• The estimand framework (ICH E9 (R1)) was introduced to standardize the definition of treatment effects.
• Early reviews indicated low adoption and lack of clarity regarding intercurrent events.

Purpose of the Study:

• To assess the current application of the estimand framework in adaptive design clinical trials.
• To identify challenges and gaps in the implementation of the estimand framework for adaptive trials.

Main Methods:

• Systematic review of phase 2-4 adaptive design RCT protocols and statistical analysis plans published in 2023.
• Searched BMJ Open and Trials journals.
• Data extraction focused on estimand specification, handling of intercurrent events, and interim analyses.

Main Results:

• Only 3% of 146 eligible articles specified a primary estimand; 2% detailed all estimand attributes.
• 64% of articles described intercurrent events, but none linked them to the primary estimand.
• No articles specified estimands for interim analyses or considered adaptation impacts.

Conclusions:

• The estimand framework is underutilized in adaptive design RCTs.
• Widespread adoption is necessary for clear reporting and interpretation of trial results.
• Specific guidance for applying the estimand framework to adaptive designs is required.