Two-stage randomized trial design for testing treatment, preference, and self-selection effects for count outcomes.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Health Services Research

Background:

• Traditional clinical trials focus on treatment effects, often overlooking patient preferences.
• Patient psychology towards treatments is crucial for optimizing patient-centered care strategies.
• Existing two-stage randomized trial designs are limited to continuous and binary outcomes.

Purpose of the Study:

• To extend the two-stage randomized trial methodology to accommodate count outcomes.
• To develop statistical methods for evaluating patient preference, selection, and treatment effects in count data.
• To provide sample size calculation formulas for these effects in two-stage trials.

Main Methods:

• Derivation of test statistics for preference, selection, and treatment effects for count outcomes.
• Development of closed-form sample size formulae for each effect.
• Conducting simulations to evaluate the performance of unstratified and stratified designs.

Main Results:

• The study successfully extends the two-stage randomized trial design to include count outcomes.
• New test statistics and sample size formulas are presented for preference, selection, and treatment effects.
• Simulation results demonstrate the properties of the proposed statistical methods.

Conclusions:

• The extended two-stage randomized trial design offers a valuable framework for analyzing count data while considering patient preferences.
• This methodology enhances the ability to conduct patient-centered research by incorporating psychological factors.
• The developed methods are applicable to real-world scenarios, such as antimicrobial use in end-of-life care.