连续重评估方法无差异区间半宽度δ的校正与毒性概率估计

Half-width (δ) of Indifferent Interval and Toxicity Probabilities Calculation in Continual Reassessment Method

本文介绍了抗癌新药研发中Ⅰ期临床试验连续重评估方法(Continual reassessment method,CRM)中无差异区间半宽度δ的概念与校正方法。利用R软件,详细阐述了半宽度δ校正的实施步骤,最后,介绍了CRM毒性概率的生成方法和模型验证,旨在为今后我国抗癌新药研发Ⅰ期临床试验设计提供方法学支持。

We introduced the concept and calculation method of half-width( δ) of indifferent interval in continual reassessment method( CRM) for phase Ⅰ clinical trial design of anti-cancer drugs. Using the R statistical software,we presented in detail about the calibration approach of half-width( δ). In the end,we introduced the approaches of toxicity probabilities calculation and model validation using R statistical software. We hope this article can provide methodological support for the design of phase Ⅰ clinical trial of anti-cancer drug for our country.