Biostatistics and Data Sciences, Boehringer Ingelheim Pharmaceuticals Inc, Ridgefield, CT, USA.
Department of Statistics and Actuarial Science, College of Mathematics and Statistics, Chongqing University, Chongqing, China.
Stat Methods Med Res. 2020 Feb;29(2):522-540. doi: 10.1177/0962280219840383. Epub 2019 Apr 8.
Traditionally, statistical methods for futility analysis are developed based on a single study. To establish a drug's effectiveness, usually at least two adequate and well-controlled studies need to demonstrate convincing evidence on its own. Therefore, in a standard clinical development program in chronic diseases, two independent studies are generally conducted for drug registration. This paper proposes a statistical method to combine interim data from two independent and similar studies for interim futility analysis and shows that the conditional power approach based on combined interim data has better operating characteristics compared to the approach based on single-trial interim data, even with small to moderate heterogeneity on the treatment effects between the two studies.
传统上，无效性分析的统计方法是基于单个研究开发的。为了确定药物的疗效，通常需要至少两项充分且对照良好的研究来单独证明其有效性。因此，在慢性病的标准临床开发计划中，通常会进行两项独立的研究来进行药物注册。本文提出了一种统计方法，用于合并来自两项独立且相似研究的中期数据，以进行中期无效性分析，并表明基于合并中期数据的条件功效方法与基于单试验中期数据的方法相比，具有更好的操作特性，即使在两项研究之间的治疗效果存在小到中度异质性的情况下也是如此。
