Sequential parallel comparison design for "gold standard" noninferiority trials with a prespecified margin

Area of Science:

• Clinical Trials Methodology
• Psychiatric Research
• Statistical Analysis

Background:

• Three-arm noninferiority trials are the "gold standard" for evaluating experimental treatments in mental disorders.
• These trials often fail to demonstrate treatment superiority due to significant placebo response rates.
• The sequential parallel comparison design (SPCD) is a potential strategy to mitigate placebo effects.

Purpose of the Study:

• To develop a hypothesis-testing method for "gold standard" noninferiority trials using the SPCD.
• To create a corresponding sample size calculation method for these trials.
• To evaluate the performance of the proposed methods in a simulation study.

Main Methods:

• Development of a novel hypothesis-testing framework for "gold standard" noninferiority trials incorporating the SPCD.
• Derivation of a sample size calculation formula tailored to the SPCD in this trial design.
• Simulation studies to assess the type I error rate, power, and sample size accuracy of the proposed methods.

Main Results:

• The proposed hypothesis-testing method successfully achieved the nominal type I error rate and desired statistical power.
• The developed sample size calculation method demonstrated adequate accuracy in power estimation.
• These findings suggest the SPCD is a viable approach for "gold standard" noninferiority trials.

Conclusions:

• The newly developed hypothesis-testing and sample size calculation methods are effective for "gold standard" noninferiority trials using the SPCD.
• This approach offers a statistically sound way to conduct noninferiority trials in mental disorders, accounting for placebo effects.
• The methods provide researchers with tools to design more reliable and informative clinical trials in psychiatry.