Wednesday, 9 June 2021

Robustness of the Exact t Method for a Three-arm Clinical Endpoint Bioequivalence Study under Non-normality| Chapter 7 | Theory and Practice of Mathematics and Computer Science Vol. 11

 A clinical endpoint bioequivalence (BE) study seeks to determine the BE of a generic drug (TEST) and an innovator drug (REF). A placebo (PLB) is usually included to demonstrate the study's sensitivity. BE is established if TEST is shown to be superior to PLB, REF is shown to be superior to PLB, and TEST is shown to be equivalent to REF. As a result, for a clinical endpoint BE study, an overall BE test consists of two superiority tests (TEST vs. PLB and REF vs. PLB) and one equivalence test (TEST vs. REF). Chang et al. [1] used the joint distribution of sample means and variances to calculate the sample size and power for an overall BE test based on one superiority test (TEST vs. PLB) and an equivalence test (TEST vs. REF) (we call this a Z-ChiSquare method). We previously proposed an exact method to calculate the power and sample size for an overall BE test based on two superiority tests (TEST vs. PLB, REF vs. PLB) and one equivalence test (TEST vs. REF) directly using a multivariate non-central t distribution (we call this the Exact-t method) for a clinical endpoint BE study with two superiority tests and one equivalence test. Yang and Sun demonstrated that the Exact-t method works. When compared to the Z-ChiSquare method, it is more computationally efficient and accurate when the sample size is small. However, under the normality assumption, these methods were generally validated by simulation. In reality, data can deviate from the norm (e.g., be skewed). We investigate the robustness of the Exact-t method and the Z-ChiSquare method when data is mildly or severely skewed in this paper. It turns out that as long as the mean and variance of the data are correctly specified, both methods remain accurate even when the data is severely skewed. One thing to keep in mind is that when the data is more skewed, the sample size required to achieve the desired power is larger. As a result, When calculating power and determining sample size for a three-arm clinical endpoint study, the Exact-t method is recommended.

Author (s) Details

Dr. Aotian Yang
The George Washington University, 2121 I Street, NW, Washington, DC, 20052, United States.

Dr. Wanjie Sun
U.S. Food and Drug Administration, 10903 New Hampshire Ave, Silver Spring, MD, 20993, United States.

View Book :  https://stm.bookpi.org/TPMCS-V11/article/view/1312

No comments:

Post a Comment