STATISTICAL MODELS IN REPLICATED CROSSOVER DESIGN IN BIOEQUIVALENCE CLINICAL TRIAL: FIXED EFFECT MODEL VERSUS RANDOM EFFECT MODEL

Sujata Suvarnapathki; Manik Khaparde

Volume : IV, Issue : VII, July - 2015

STATISTICAL MODELS IN REPLICATED CROSSOVER DESIGN IN BIOEQUIVALENCE CLINICAL TRIAL: FIXED EFFECT MODEL VERSUS RANDOM EFFECT MODEL

Sujata Suvarnapathki, Manik Khaparde

Abstract :

In Analysis of Variance, there are two types of factors fixed effect and random effect. An appropriate type of effect is chosen depending on the context of the problem. In Fixed effect modelling, the interest lies in comparison of the specific levels e.g. Formulation or treatment for a particular drug product. In random effect modelling, the levels of the factor are drawn randomly from universe. In case of Bioequivalence studies, EMEA and US FDA differ on the view of treating subject effect. EMEA assumes subject effect as fixed whereas US FDA assumes it to be random. In this paper, three methods for statistical analysis of Pharmacokinetic parameters are compared for random effect and fixed effect model for replicated crossover design in a bioequivalence study. Conclusions: The results of the statistical analyses indicate that 90% confidence interval for the three approaches are not same. In general, Method C gives wider confidence intervals.