治疗-对照差临床意义判别所需样本量的测定

Sample Size Determination for Discriminating the Clinical Significance of Treatment-Control Difference

目的 提出用于治疗 -对照差临床意义判别的样本量测定方法。方法 依据治疗 -对照差与最小临床承认疗效差量的比较推导出所需样本量测定公式 ,以MonteCarlo方法展示其行为。结果 当最小临床承认疗效差量取值为零时 ,本文方法还原为渐近正态法。由所测样本量产生的观测功效与预定功效吻合。结论 对于以判别治疗 -对照差临床意义为目的的临床试验 ,可用本文方法测定所需样本量。

Objective This paper proposes a procedure of determining the required sample sizes for discriminating the clinical significance of treatment control difference. Methods The equation of determination is derived from comparing the treatment control difference with the minimal clinically relevant therapeutic difference. Its performance is demonstrated by the Monte Carlo method. Results It reduces to the asymptotic normal procedure when the value is taken as zero for the minimal clinically relevant therapeutic difference. The observed power of the test coincides with the prescribed power when the sample sizes determined by the procedures are employed. Conclusion The procedure can be applied to determining the required sample sizes for the clinical trials with the purpose of discriminating the clinical significance of treatment control difference.