Cassie Dong Xiaoyu, Bian Yuanyuan, Tsong Yi, Wang Tianhua
a Global Biostatistical Science, Amgen Inc. , Washington , DC , USA.
b ORISE Intern 2016, Department of Statistics , University of Missouri-Columbia , Columbia , Missouri , USA.
J Biopharm Stat. 2017;27(2):317-330. doi: 10.1080/10543406.2016.1265546. Epub 2017 Jan 5.
The equivalence test has a wide range of applications in pharmaceutical statistics which we need to test for the similarity between two groups. In recent years, the equivalence test has been used in assessing the analytical similarity between a proposed biosimilar product and a reference product. More specifically, the mean values of the two products for a given quality attribute are compared against an equivalence margin in the form of ±f × σ, where ± f × σ R is a function of the reference variability. In practice, this margin is unknown and is estimated from the sample as ±f × S. If we use this estimated margin with the classic t-test statistic on the equivalence test for the means, both Type I and Type II error rates may inflate. To resolve this issue, we develop an exact-based test method and compare this method with other proposed methods, such as the Wald test, the constrained Wald test, and the Generalized Pivotal Quantity (GPQ) in terms of Type I error rate and power. Application of those methods on data analysis is also provided in this paper. This work focuses on the development and discussion of the general statistical methodology and is not limited to the application of analytical similarity.
等效性检验在药物统计学中有广泛应用，我们需要检验两组之间的相似性。近年来，等效性检验已用于评估拟议的生物类似药产品与参比产品之间的分析相似性。更具体地说，将两种产品对于给定质量属性的均值与以±f×σ形式表示的等效界值进行比较，其中±f×σR是参比变异性的函数。在实际中，该界值未知，需从样本中估计为±f×S。如果我们在均值的等效性检验中使用这个估计的界值和经典t检验统计量，Ⅰ类错误率和Ⅱ类错误率可能都会膨胀。为解决这个问题，我们开发了一种基于精确值的检验方法，并将该方法与其他提出的方法（如Wald检验、受限Wald检验和广义枢轴量（GPQ））在Ⅰ类错误率和检验效能方面进行比较。本文还提供了这些方法在数据分析中的应用。这项工作侧重于通用统计方法的开发和讨论，并不局限于分析相似性的应用。
