【摘要】 目的:介绍临床随访研究中生存分析资料的log-rank检验所需样本含量的估计法。方法:以离散性Markov链拟合生存过程,据此计算log-rank统计量的数学期望和方差,导出样本含量估计公式。结果:实例分析表明,该法能较好反映实际情况,应用灵活。结论:本法是一种有效、可行的样本含量估计法,值得推荐。更多还原

【Abstract】 Objective:To introduce method for estimation of sample size needed in log-rank test of survival data in clinical follow-up study. Methods: The survival process was modeled by discrete Markov chain and the asymptotic expectation and variance of log-rank statistics were calculated thereafter. The sample size formula was thus derived. Results: Practical case analysis indicated that actual situation may be appropriately reflected by the method and be conveniently used. Conclusion: The method for estimation of sample size described above is effective and practical and worth to be recommended.更多还原