论文部分内容阅读
为论述生存研究样本量测定的基本原理和方法.将用于比例检验样本量测定的经典渐近正态法加以扩展以适应终检形成广义渐近正态法。结果显示，与已有校正终检样本量测定方法相比．该方法的特点是与生存率检验相匹配，摆脱了指数分布的假设，在无终检时还原为经典方法。该方法适用于癌症生存研究的设计。
To discuss the basic principles and methods for determining the sample size for survival studies, the classical asymptotic normal method for the determination of sample sizes for scaling tests is extended to accommodate the final check to form generalized asymptotic normal methods. The results show that, compared with the method of determining the final calibration sample size. The method is characterized by matching with the survival rate test, getting rid of the hypothesis of exponential distribution and returning to the classic method without final inspection. This method is suitable for the design of cancer survival research.