Evaluating treatment efficacy by multiple end points in phase II acute heart failure clinical trials: analyzing data...

Area of Science:

• Cardiology
• Clinical Trials
• Biostatistics

Background:

• Acute heart failure (AHF) studies require robust statistical methods to evaluate interventions across multiple clinical endpoints simultaneously.
• Traditional methods may lack the power to detect significant treatment effects when considering several outcomes at once.

Purpose of the Study:

• To assess the concomitant, simultaneous effects of interventions on multiple endpoints in Phase II acute heart failure studies using global statistical methods.
• To compare the statistical power of different methods for evaluating treatment effects on various AHF outcomes.

Main Methods:

• Simulations were used to evaluate statistical methods for assessing intervention effects on dyspnea relief (2 measures), hospital stay, heart failure worsening, mortality, and readmission.
• Scenarios included large and very large relative improvements, with typical placebo responses and correlations among endpoints.

Main Results:

• The average Z score method demonstrated high power (>70% with ≥75 patients/group) for detecting a 35% relative improvement across six endpoints.
• Analyzing only dyspnea relief showed significantly lower power compared to the average Z score approach.
• Other evaluated methods generally yielded lower statistical power than the average Z score.

Conclusions:

• The average Z score method allows for the detection of therapeutic efficacy in AHF studies with sample sizes of 100-150 patients per group.
• This global statistical approach offers approximately double the power compared to assessing dyspnea alone, facilitating more efficient clinical trial design.