Moussa M A
Department of Community Medicine, Faculty of Medicine & Health Sciences, Al-Ain, United Arab Emirates.
Comput Biol Med. 1993 Jan;23(1):69-76. doi: 10.1016/0010-4825(93)90109-e.
The paper deals with potency ratio estimation of parallel curve analytic dilution assays in case log dose-response relationship could be reasonably described by a parabola. It comprises: (1) testing the adequacy and validity of the quadratic model by the analysis of variance; and (2) estimation of the relative potency of an unknown in relation to a standard preparation, its standard deviation and fiducial limits for its true value. The method is applicable whenever successive doses of an unknown are a constant multiple of a standard in randomized blocks or completely randomized design. The method can be generalized to polynomial models of higher order.
本文探讨在对数剂量-反应关系可用抛物线合理描述的情况下，平行曲线分析稀释试验的效价比估计。它包括：（1）通过方差分析检验二次模型的充分性和有效性；（2）估计未知物相对于标准制剂的相对效价、其标准差及其真值的置信限。该方法适用于在随机区组或完全随机设计中，未知物的连续剂量是标准物的恒定倍数的任何情况。该方法可推广到更高阶的多项式模型。
