序贯安慰剂平行对照设计和双向富集设计的样本量估计

摘要

目的 探讨临床试验中序贯安慰剂平行对照设计和双向富集设计的样本量计算方法.方法 在不同参数设置条件下,计算序贯安慰剂平行对照设计和双向富集设计所需要的样本量,并与传统的平行组设计、交叉设计等进行比较.结果 序贯安慰剂平行对照设计和双向富集设计相比于传统的平行组设计、交叉设计需要的样本量更少,其中,双向富集设计的优势更为明显.对于连续性结局变量,第二阶段的疗效越大,第一阶段的权重越小,需要的样本量越小.对于二分类结局变量,得分检验参数取9.8和6.2时,SPCD和TED分别需要的样本量最少.假定两阶段疗效相同时,第一阶段分配到安慰剂组的比例取0.57和0.50时,SPCD和TED需要的样本量最少.结论 在安慰剂效应较高或是没有彻底治愈方案的慢性疾病试验中,序贯安慰剂平行对照设计和双向富集设计相比于传统设计有很大的优势.%Objective Our purpose is to explore and discuss the problem of sample size estimation while applying the sequential parallel comparison design and the two-way enriched design to clinical trials. Methods Calculating the sample size required for the sequential parallel comparison design and the two-way enriched design under different parameter setting conditions and comparing the result to conventional designs. Results Comparing to traditional designs, smaller sample sizes are needed for the sequential parallel comparison design and the two-way enriched design, among which, the two-way enriched design is more preferable. As to continuous outcome variable, when larger treatment effect is set for the second stage and lower weight is chose for the first stage, smaller total sample size is required. As to dichotomous outcomes, when the score test parameter reaches 9.8 and 6.2, smallest sample sizes are needed for SPCD and TED, respectively. Least numbers of subjects will be required for SPCD and TED, separately, when we choose the rate of placebo group during the first stage to be 0.57 and 0.50, under the hypothesis that the treatment effects for both stages are equal. Conclusion In the trials of chronic diseases with relatively high placebo response rate or limited cure schemes, the sequential parallel comparison design and the two-way enriched design are more advantageous compared with traditional designs.