中文摘要：

目的针对两组计量资料的非劣效性评价问题，给出一种新的统计检验方法。方法根据计量资料的不同评价分界值，给出相应的试验组和对照组的“有效率”，分别以两组的“有效率”为横轴和纵轴，连成P-P曲线并计算曲线下面积，然后根据这一统计量进行统计检验，并通过模拟实验对检验效能进行研究。结果给出了P-P曲线方法的检验公式，使用这种方法更易于选择非劣效界值，而且对数据的分布没有任何限制，在偏态情况下比传统方法有更高的检验效能。结论该方法适合用于Ⅱ期探索性临床试验研究，具有应用和研究价值。

英文摘要：

Objective In this study, we presented a new statistical method to assess the non-inferiority for two-ann quantitative outcome. Methods Efficacy rates from the test drug and positive control group were obtained according to the thresholds formed a probability and probability （P-P） curve ,where the horizontal axis is the efficacy rate from the positive con- trol drug and the vertical axis is one from the test drug. We calculated the area under the P-P curve and performed the statistical test with this new-devised statistics. In addition, a series of simulation studies were performed to test the statistical power for this method. Results We presented a test formula for P-P curve method, which is easier to choose the non-inferiority margin. Fur- thermore, this method has no limitation for the data distribution, especially in the case of skewed distribution, it could provide high statistical power for this method. Conclusion This method could be effective in the explorative study and it deserves prac- tical application and further studies.