Who Are We Missing?: A Principled Approach to Characterizing the Underrepresented Population

Area of Science:

• Clinical Epidemiology
• Biostatistics
• Health Services Research

Background:

• Randomized controlled trials (RCTs) are crucial for causal inference but face challenges in generalizability due to subgroup heterogeneity and underrepresentation.
• Extending RCT findings to broader target populations requires methods to address these limitations.

Purpose of the Study:

• To develop and validate a novel framework for identifying and characterizing underrepresented subgroups within RCTs.
• To enhance the generalizability of RCT findings to real-world populations.

Main Methods:

• Introduction of the Rashomon Set of Optimal Trees (ROOT), an optimization-based approach to characterize underrepresented groups.
• ROOT minimizes the variance of the target average treatment effect estimate to improve precision.
• Application of the methodology to extend inferences from the START trial to the TEDS-A population.

Main Results:

• ROOT effectively characterizes underrepresented populations with interpretable features.
• The approach demonstrated improved precision and interpretability compared to existing methods in synthetic data experiments.
• Successful application to refine target populations and enhance generalizability from a clinical trial.

Conclusions:

• The proposed ROOT framework provides a systematic approach to refine target populations in RCTs.
• This method enhances decision-making accuracy and aids in informing future research in diverse populations.
• ROOT facilitates more precise and interpretable treatment effect estimations, improving the external validity of clinical trial findings.