Residuals and outliers in replicate design crossover studies

Area of Science:

• Pharmacokinetics and Drug Development
• Biostatistics
• Regulatory Science

Background:

• Outliers in bioequivalence trials pose interpretation and handling challenges.
• Standard 2x2 crossover studies cannot statistically differentiate outlier types.
• Regulatory guidelines differ on outlier exclusion based on their origin.

Purpose of the Study:

• To propose a method for distinguishing and classifying outliers in replicate crossover bioequivalence trials.
• To provide a simple yet effective diagnostic tool for outlier identification.
• To aid in appropriate data handling for bioequivalence analysis.

Main Methods:

• Utilizing replicate design (2-treatment, 4-period) crossover studies.
• Calculating and plotting three types of residuals derived from orthogonal contrasts.
• Correlating residual types with subject outliers, subject-by-formulation outliers, and single-data-point outliers.

Main Results:

• The proposed residual analysis effectively distinguishes three types of outliers.
• Subject outliers are generally unproblematic for bioequivalence analysis.
• Subject-by-formulation outliers require careful consideration; single-data-point outliers may be removable.

Conclusions:

• A novel residual-based diagnostic tool facilitates outlier classification in replicate crossover designs.
• This method supports appropriate statistical handling of outliers in bioequivalence studies.
• Improved outlier management enhances the reliability of bioequivalence trial outcomes.