Abstract

An initial proposal was made to start 30 monkeys in the run-in period of a preclinical translational research study, to have 24 or more animals qualify for randomization in the subsequent treatment period. Based upon data from previous studies, Bayesian posterior prediction indicated that successful enrolment was highly unlikely. At least 67 animals were required to achieve an acceptable posterior predictive probability of success. Importantly, we leveraged these feasibility analyses to introduce our preclinical scientist collaborators to a Bayesian strategy for probability-based decision making. We provided them with a generous helping of graphics to effectively and efficiently illustrate Bayesian concepts and methods. We present our 4P strategy for collaboration with preclinical scientists: patience, persistence, positioning, and privilege. We discuss the alignment of the Bayesian and 4P strategies with goals common to pharmaceutical researchers: scientific innovation; stochastic intelligence and statistical literacy of team members; team collaboration and collegial partnerships; ethical acuity; and fiscal stewardship. Our article is as much about successfully reaching out to preclinical scientists, and introducing them to the Bayesian strategy, as it is about that strategy successfully addressing the animal enrolment question. This article is written at a statistical level accessible to both preclinical scientists and statisticians.

摘要

最初的提议是在临床前转化研究的磨合期开始 30 只猴子，在随后的治疗期有 24 只或更多动物有资格进行随机化。根据先前研究的数据，贝叶斯后验预测表明成功注册的可能性很小。至少需要 67 只动物才能达到可接受的成功后验预测概率。重要的是，我们利用这些可行性分析向我们的临床前科学家合作者介绍了基于概率的决策的贝叶斯策略。我们为他们提供了大量的图形帮助，以有效和高效地说明贝叶斯概念和方法。我们提出了与临床前科学家合作的 4P 策略：耐心、坚持、定位和特权。我们讨论了贝叶斯和 4P 策略与制药研究人员共同目标的一致性：科学创新；团队成员的随机情报和统计素养；团队合作和学院合作；道德敏锐度；和财政管理。我们的文章是关于成功接触临床前科学家，并将他们介绍给贝叶斯策略，因为它是关于成功解决动物登记问题的策略。本文是在临床前科学家和统计学家都可以访问的统计水平上编写的。和财政管理。我们的文章是关于成功接触临床前科学家，并将他们介绍给贝叶斯策略，因为它是关于成功解决动物登记问题的策略。本文是在临床前科学家和统计学家都可以访问的统计水平上编写的。和财政管理。我们的文章是关于成功接触临床前科学家，并将他们介绍给贝叶斯策略，因为它是关于成功解决动物登记问题的策略。本文是在临床前科学家和统计学家都可以访问的统计水平上编写的。