Inference of Bioequivalence for Log-Normal Distributed Data with Unspecified Variances

• Activity Number: 421
• Type: Topic Contributed
• Date/Time: Tuesday, August 6, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Siyan Xu*+ and Steven Hua and Ronald Menton and Kerry Barker and Sandeep Menon and Ralph D'Agostino, Sr. and Mo Pei
• Companies: Boston University and Pfizer Research and Pfizer Inc. and Pfizer Inc. and Pfizer Inc. and Boston University and Boston University
• Keywords: Bioequivalence ; Likelihood function ; unspecified variances ; Bayesian model ; TOST

Abstract

The distribution of PK parameters is often assumed to be log-normal, thus bioequivalence (BE) is usually assessed in the difference of a logarithmically transformed PK parameter (d). In the presence of unspecified variances, test procedures like two one-sided tests (TOST) and profile likelihood replace them with other estimators, and Bayesian model integrates them out from the posterior distribution. Those methods limit our knowledge on the extent that inference about BE is affected by the variability of PK parameters. In this paper, we propose a likelihood approach that retains the unspecified variances in the model and partitions the entire likelihood function into two components: F-statistic function for variances and t-statistic function for d. Demonstrated with published real life data, the proposed method not only produces results same as TOST and comparable to results using Bayesian method, it also helps identify a range of values of the unspecified variances where BE is more likely to be achieved. The second advantage is missing in TOST and Bayesian method, suggesting the effectiveness in making inference about the extent that BE is affected by the unspecified variances.