某一種新藥要成為市場上參考藥的合格替代藥必須經過非劣性檢定。目前文獻中針對右設限存活資料，是根據存活函數差異進行的非劣性檢定。本文則根據限制平均壽命差異提出兩種非劣性檢定，其中一種檢定是根據在某些既定時間點的限制平均壽命差異，另一種則是根據在上述時間點的最大限制平均壽命差異。本文藉由模擬研究探討所提兩種檢定方法的檢定之型I誤差率及檢定力的表現。最後，以一則實例說明上述兩種非劣性檢定之應用。 The noninferiority test is required for approving a new drug which provides with an alternative to the reference drug in the market. The current noninferiority tests for randomly right-censored data are based on the difference between two survival functions. In this article, we proposed two noninferiority tests based on the difference between two restricted mean lifetimes. One is constructed according to the differences between the restricted mean lifetimes at some specific time points, the other is based on the maximum of all the possible differences over the specific time points. A simulation study is then conducted to compare the type I error rate and power of the proposed tests. Finally, the application of the proposed methods is illustrated by using a real data.
