Biswas A
Applied Statistics Unit, Indian Statistical Institute, 203 B.T. Road, Kolkata 700 035, India.
Stat Methods Med Res. 2001 Oct;10(5):353-64. doi: 10.1177/096228020101000504.
In phase III clinical trials the problem is to compare two treatments or therapies for their effects. The randomized 50:50 allocation to the two competing treatments is popular in most of the cases. In contrast there is a class of adaptive designs available in literature for this purpose. The object is to treat a larger number of patients by the eventual better treatment. A few such adaptive designs were conducted out in practical applications and consequently a lot of controversy arose. In spite of that there is steady theoretical progress in this area over the past three decades. The present paper reviews some of the real-life adaptive clinical trials and discusses some logistics of such trials.
在III期临床试验中，问题在于比较两种治疗方法或疗法的效果。在大多数情况下，将患者以50:50的比例随机分配到两种相互竞争的治疗方法中很常见。相比之下，文献中有一种适用于此目的的适应性设计。目的是通过最终更好的治疗方法治疗更多患者。一些这样的适应性设计在实际应用中进行了，结果引发了很多争议。尽管如此，在过去三十年里，这一领域仍取得了稳步的理论进展。本文回顾了一些实际的适应性临床试验，并讨论了此类试验的一些后勤问题。
