Finkelstein D M, Schoenfeld D A
Biostatistics Department, Harvard School of Public Health, Boston, MA 02115.
Stat Med. 1994 Sep 15;13(17):1747-54. doi: 10.1002/sim.4780131706.
A primary endpoint of AIDS trials is the length of survival. Often there is auxiliary information available on measures of disease progression, which significantly alter the risk of mortality. This paper explores the use of this information in obtaining a refined estimate of survival, and a test based on this estimate. The methods are applied to an AIDS clinical trial, and results of simulations are provided which compare the approach to standard methods for survival analysis.
艾滋病试验的主要终点是生存时长。通常，关于疾病进展的测量存在辅助信息，这些信息会显著改变死亡风险。本文探讨了如何利用这些信息来获得生存的精确估计值，以及基于该估计值的检验。这些方法应用于一项艾滋病临床试验，并提供了模拟结果，将该方法与生存分析的标准方法进行了比较。
