基于泊松分布的两种治疗方法的非劣性评价

Non-inferiority Evaluation on the Two Therapies Based on Poisson Distribution

针对在样本量不等的情况下基于泊松分布的两种治疗方法的非劣性评价问题,考虑了该问题下的Score检验、Wald检验以及对数变换的几种不同的检验方法,并以一个来自于乳腺癌研究的实际例子考察了以上这几种统计量的经验表现。

Aiming at problems of non-inferiority evaluation on the two therapies based on Poisson distribution under different amount of samples, several different testing methods as Score test, Wald test and some other logarithm-transposition tests were thought over. The experimental performance of the above statistics was studied according to a real case of breast cancer.