Update on the statistical analysis of bioequivalence studies

Area of Science:

• Pharmacokinetics and Biopharmaceutics
• Statistical Methods in Drug Development
• Regulatory Science

Background:

• Assessing bioequivalence between test and reference drug formulations is crucial for regulatory approval.
• Understanding the distribution of bioequivalence characteristics and associated consumer risks is essential.
• Existing methods require careful consideration of both consumer and producer risks.

Purpose of the Study:

• To review statistical methods for assessing drug bioequivalence.
• To identify a method that minimizes consumer risk while maintaining an acceptable producer risk.
• To recommend a practical strategy for bioequivalence decisions.

Main Methods:

• Evaluation of statistical procedures for bioequivalence assessment.
• Focus on 90%-confidence intervals for the ratio of expected medians.
• Consideration of both parametric (logarithmic normal distribution) and nonparametric (distribution-free) approaches.
• Analysis of bioequivalence range modifications for different pharmacokinetic parameters.

Main Results:

• A recommended strategy involves using the shortest 90%-confidence interval for the ratio of expected medians within the bioequivalence range (80-120%) for rate and extent of absorption.
• The nonparametric 90%-confidence interval is the preferred method when the logarithmic normal distribution assumption is violated.
• A decision rule for time to maximum concentration (tmax) is also presented.

Conclusions:

• The proposed statistical strategy effectively balances consumer and producer risks in bioequivalence assessment.
• The method is adaptable to different pharmacokinetic parameters and distributional assumptions.
• This approach supports robust decision-making in drug formulation bioequivalence studies.