Assessment of equivalence on multiple endpoints

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmaceutical Research

Background:

• Clinical trials often assess therapeutic equivalence across multiple primary endpoints.
• Osteoarthritis treatment studies exemplify this, using endpoints like investigator/patient global assessments and pain.
• The conventional intersection-union test for simultaneous equivalence can be overly conservative.

Purpose of the Study:

• To explore simultaneous equivalence assessment for multiple endpoints in clinical trials.
• To investigate the conservativeness of the intersection-union test through simulation and computation.
• To propose a novel method for lowering confidence levels while maintaining Type I error rates for independent, normally distributed endpoints.

Main Methods:

• Simulation and computational analysis to evaluate the conservativeness of existing methods.
• Development of a new statistical method for simultaneous equivalence assessment.
• Introduction of a step-up procedure for selecting endpoints demonstrating equivalence, controlling the experimentwise error rate.

Main Results:

• The conventional intersection-union test demonstrates conservativeness in simultaneous equivalence assessments.
• A new method is presented that allows for reduced confidence levels without compromising Type I error rates under specific conditions (normal, independent distributions).
• The step-up procedure effectively identifies multiple endpoints for demonstrated equivalence while controlling the overall error rate.

Conclusions:

• The study provides insights into the conservativeness of traditional methods for multiple endpoint equivalence testing.
• Novel statistical approaches are offered to enhance the efficiency and accuracy of therapeutic equivalence trials.
• The proposed step-up procedure offers a practical solution for demonstrating equivalence across multiple endpoints in complex trial designs.