生存研究样本量设计表及其使用方法

Design tables of the required sample sizes for survival studies and their appliocations

目的 :本文就两样本生存研究所需样本量用广义渐近正态法设计出 15 30个方案 ,制成设计表供临床使用。方法 :利用C语言编制程序 ,按通用格式打印设计表。结果 :常用方案查表可得。这些与两样本生存率检验相匹配 ,具有还原性 ,摆脱了比例危险的假设 ,在常见临床条件下观测功效符合预定功效。结论 :这些设计表适用于两样本癌症临床研究的设计 。

Objective: Design tables of the required sample sizes for survival studies including 1,530 projects were made out based on the generalized asymptotic normal procedure for clinical applications. Methods: The procedure was programmed in the C language and the output was printed in the destgn tables with usual form. Results: Common projects can be found in the tables. They are matched with the test on two survival rates, reducible when there is no censoring, free from the assumption of proportional hazards, The observed power coinccides with the prescribed power in usual clinical situations. Conclusion: The tables are appropriate with planning clinical cancer studies, especially useful for clinicians or the relevant investigators.