Priors and decision thresholds in phase 2 and phase 3 randomized controlled trials evaluating drug efficacy using...

Area of Science:

• Biostatistics
• Clinical Trials
• Pharmacology

Background:

• Bayesian methods are increasingly utilized in clinical trials for drug efficacy evaluation.
• These methods offer flexibility by incorporating prior knowledge into statistical analyses.
• However, challenges in interpretation exist due to a lack of standardized reporting.

Purpose of the Study:

• To systematically review and describe the application of priors and decision thresholds in Phase 2 and 3 randomized controlled trials (RCTs) using Bayesian methods for drug efficacy assessment.
• To identify current practices and variations in the reporting of Bayesian approaches in clinical research.

Main Methods:

• A systematic review of Phase 2 and 3 RCTs evaluating drug efficacy via Bayesian inference was performed.
• Databases searched included MEDLINE, EMBASE, and Cochrane up to September 2022.
• Data extracted focused on prior type, parameters, justification, sensitivity analysis, and the presence and value of a priori decision thresholds.

Main Results:

• Out of 1161 screened articles, 69 trials (91 comparisons) were included.
• Priors were assigned to treatment effect in 51% of cases and to each arm in 37%; prior distribution details were provided in 59%.
• A decision threshold was a priori defined in 68% of studies, with values ranging from 70% to 99% (median 95%).

Conclusions:

• Inconsistent description of priors and heterogeneous use of decision thresholds highlight a need for standardized guidelines.
• Clearer standards for the application and reporting of Bayesian methods in clinical trials are essential.
• Improved transparency and consistency in reporting Bayesian trial results are required.