The analysis of failure time data in crossover studies

Area of Science:

• Cardiology
• Clinical Trials
• Biostatistics

Background:

• Exercise tests are crucial for evaluating angina pectoris drug efficacy in clinical trials.
• Crossover designs are frequently used, but current analysis methods for exercise times are insensitive and biased.
• Handling censored data in exercise test analysis requires robust statistical approaches.

Purpose of the Study:

• To introduce survival analysis as a superior method for analyzing exercise test data in angina pectoris crossover trials.
• To demonstrate how survival analysis overcomes the limitations of traditional methods, offering increased sensitivity and reduced bias.
• To highlight the practical clinical relevance of using median survival times for interpreting treatment effects.

Main Methods:

• Adaptation of survival analysis techniques for crossover trial designs.
• Application of the Cox proportional hazards model for advanced analysis.
• Illustration using a two-period crossover trial comparing atenolol with atenolol plus nifedipine.

Main Results:

• Survival analysis provides more sensitive and less biased results compared to conventional methods.
• Median survival times offer a clinically interpretable measure of treatment efficacy.
• The methodology is adaptable to both two-period and three-period crossover designs.

Conclusions:

• Survival analysis is a powerful and appropriate statistical tool for analyzing exercise test data in angina pectoris crossover trials.
• This method enhances the reliability and interpretability of clinical trial results.
• The use of median survival times aids in understanding the clinical significance of therapeutic interventions.