Simon R
Biometrics. 1981 Dec;37(4):723-31.
This paper proposes a new design for the augmentation of a randomized clinical trial through the participation of private physicians who do not randomize individual patients. Each private physician is randomly assigned a study treatment which he or she uses consistently for all eligible patients. A statistically rigorous method of analysis is described, the efficiency of this design is evaluated, and its limitations discussed.
本文提出了一种新的设计，通过让不参与单个患者随机分组的私人医生参与，来扩大随机临床试验。每位私人医生被随机分配一种研究治疗方法，他或她会对所有符合条件的患者持续使用该方法。文中描述了一种统计严谨的分析方法，评估了该设计的效率，并讨论了其局限性。
