Using Trial and Observational Data to Assess Effectiveness: Trial Emulation, Transportability, Benchmarking, and...

Area of Science:

• Epidemiology
• Biostatistics
• Health Services Research

Background:

• Randomized trials and observational studies often yield conflicting results for similar research questions.
• Lack of consensus exists on when to compare these study types and how to integrate their findings for treatment effectiveness.
• Observational analyses can potentially complement randomized trial data in assessing treatment effectiveness.

Purpose of the Study:

• To explore methods for using observational analyses to complement randomized trial analyses in treatment effectiveness assessment.
• To review frameworks for designing observational studies that emulate target trials.
• To discuss approaches for estimating treatment effects in target populations using emulation and transportability.

Main Methods:

• Reviewing the framework for designing observational analyses that emulate target trials.
• Examining methods for estimating average treatment effects using observational data and transportability analyses.
• Explaining benchmarking to assess the validity of observational analyses by comparing them to trial results.
• Illustrating approaches with a pragmatic trial and its registry data emulation.

Main Results:

• Observational analyses can be designed to emulate target trials, providing a framework for comparison.
• Transportability analyses can extend trial inferences to target populations.
• Benchmarking allows for the assessment of observational analysis validity and potential joint analysis of trial and observational data.

Conclusions:

• Synthesizing trial and observational data through transportability, benchmarking, or joint analyses leverages complementary strengths.
• This integration enhances learning about comparative effectiveness.
• Combining quantitative methods with epidemiologic judgment is crucial for robust conclusions.