治疗—对照差的临床意义及其判别方法

Clinical Significance of TreatmentControl Difference and its Discriminating Approach

目的 说明治疗—对照差的临床意义及其判别方法 ,为此提出一种非虚假设检验。方法 以同源有效样本量取代样本量 ,将用于比较两率差和最小临床承认疗效差量的经典非虚假设检验加以扩展以覆盖终检。结果 这种检验可用于实现这种判别。在无终检时 ,它还原为经典检验。结论 与通常的药物疗效评价方法相比 ,这显然提高了一步。

Objective To explain the clinical significance of the treatmentcontrol difference and its discriminating approach and propose a test of non-null hypothesis for it. Methods The classical test of non-null hypothesis for comparing the difference of two rates with the minimal clinically relevant the rapeutic difference is generalized so as to cover censorship by means of using the homogenetic effective sample size in place of sample size. Results The test can be used for realizing the discrimination. It reduces to its classical counterpart when there is no censoring. Conclusion This is a remarkable progress by contrast with the usual methods in the clinical evaluation of drugs.