非虚假设IV期临床试验的设计与分析

目的:论述非虚假设IV期临床试验的设计与分析。方法:按治疗-对照差与最小可识别差量的比较建立非虚假设。将临床试验的每个中心看作一个层,以层样本分数为权作加权平均得综合治疗-对照差及其期望和方差。由此构造非虚假设分层设计基本关系式,进而推导出非虚假设IV期临床试验所需样本量和检验统计量。以M on te C arlo方法展示其行为。结果:当最小可识别差量取零时,它还原为传统的虚假设IV期临床试验所需样本量和检验统计量,其观测功效与预定功效吻合。结论:这种临床试验可用于建立试药对于有效对照的临床优效性或非劣效性。

Design and Analysis of Phase IV Clinical Trial under Non-Null Hypothesis

Objective: This paper concerns the design and analysis of phase IV clinical trial under non-null hypothesis.Methods: Comparing the treatmentcontrol difference with the minimal detectable difference leads to establish the non-null hypothesis.Each center in a clinical trial is regarded as a stratum. Taking the stratum sample fraction as a weight and making weighted average yield the summary treatment-control difference and its expectation and variance.Hence we construct the equations of basic relationship for stratified designs under non-null hypothesis.It is further used to derive the required sample size and test statistic for phase IV clinical trial under non-null hypothesis.Their performance is exhibited by Monte Carlo method.Results: They reduce to the formulae for sample size and the test statistic of the classical phase IV clinical trial under null hypothesis setting the minimal detectable difference equal to zero.The observed power coincides with the prescribed power.Conclusion: Such a clinical trial can be applied to establishing the clinical superiority or non-inferiority of tested drug versus active control.A worked example illustrates the methodology.