正态结局临床试验设计中的保证方法：方差不确定性的引出
统计方法学
2019-05-30 v2
摘要
保证方法在临床试验规划中日益流行。该方法包括引出治疗效应的先验分布，然后计算所提议试验产生“成功”结局的概率。对于正态分布的观测值，还需考虑正态方差的不确定性，但文献中如何引出方差参数的分布鲜有指导。我们提出一种简单的引出方法，并说明如何将引出的分布纳入保证计算。我们还考虑多阶段试验，其中进行更大试验的决策将取决于较小试验的结局；我们阐明了引出分布在评估所提议较小试验所提供信息中的作用。有免费软件可实现我们的方法。
引用
@article{arxiv.1702.00978,
title = {Assurance for clinical trial design with normally distributed outcomes: eliciting uncertainty about variances},
author = {Ziyad A. Alhussain and Jeremy E. Oakley},
journal= {arXiv preprint arXiv:1702.00978},
year = {2019}
}
备注
The previous version of this paper had the title "Eliciting judgements about uncertain population means and variances". The elicitation methods in this paper are unchanged from the previous version. The motivation for the methods has changed, and the scope of the paper has been widened, to show the application of the elicitation methods within the assurance method for clinical trial planning