Economizing comparative Poisson clinical trials of multiple experimental treatments by testing against the same...

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Experimental Design

Background:

• Comparative Poisson trials (Design A) typically condition on total events for inference using binomial distribution.
• Design C was recently introduced for comparing K experimental treatments to a single control, using negative multinomial distribution for inference.

Purpose of the Study:

• To compare the sample size efficiency of Design C versus K separate Design A trials.
• To evaluate the designs under uncurtailed and curtailed settings.

Main Methods:

• Simulated various combinations of Type 1 error, power, and incidence rate ratios.
• Compared expected number of subjects enrolled for Design C and Design A under different hypotheses and assumptions.

Main Results:

• Design C frequently demonstrates significant sample size savings compared to Design A.
• Savings were observed under both uncurtailed and curtailed trial settings.

Conclusions:

• Conducting a single Design C trial is often more sample size efficient than multiple Design A trials.
• Design C represents a potentially advantageous strategy for comparative clinical trials with multiple experimental arms.