Survival analysis: caveats and pitfalls
Survival analysis: caveats and pitfalls
复制标题
DOI：
10.1053/ejso.1998.0650
复制
发表时间：
1999-06-01
期刊：
影响因子：
--
通讯作者：
Murthy, NS
中科院分区：
文献类型：
--
作者：
Mathew, A;Pandey, M;Murthy, NS
背景：临床研究中的生存分析对于评估特定治疗的有效性和了解各种疾病特征的影响非常重要。有许多方法可以估计存活率及其标准误差。然而，人们不能确定这些方法是否得到了适当的处理。计算机的广泛使用使得在没有专家指导的情况下进行生存分析成为可能，但使用不适当的方法可能会产生错误的结论。大多数生物医学期刊现在建议，除了获得专家审稿人对文章的评论外，还应该由一位经验丰富的生物统计学家对每篇论文进行统计审查。这个问题在第三世界国家的论文中更加复杂，因为这些国家的所有机构可能都没有生物统计学家来指导临床医生选择适当的技术。方法：本论文涉及各种生存分析技术及其解释，使用自1982年以来Trivandrum地区癌症中心治疗的恶性上气消化道黑色素瘤患者的模态数据集。结果：Kaplan-Meier法是最适合进行生存分析的方法。中位生存时间是一种比平均值更好的汇总数据的方法。Rothman估计置信限的方法优于Peto方法，因为Peto方法计算的生存概率置信限往往超出0-1.0的范围，特别是在样本量较小的情况下。结论：本研究结果提示应采用Kaplan-Meier法进行生存分析。应尽可能提供中位生存时间，而不是依赖于平均生存时间。应将置信限作为可变性的度量来计算。一个合适的等级检验应该用来比较两个或更多的生存曲线，而不是z检验。分层分析和Cox模型，当分层分析失败时，可用于确定预后因素对生存的影响。
Background: Survival analysis in clinical studies is important to assess the effectiveness of a given treatment and to understand the effect of various disease characteristics. A number of methods exist to estimate the survival rate and its standard error. However, one cannot be certain that these methods have been handled appropriately. The widespread use of computers has made it possible to carry out survival analysis without expert guidance, but using inappropriate methods can give rise to erroneous conclusions. The majority of the biomedical journals now recommend that a statistical review of each manuscript should be carried out by an experienced bio-statistician, in addition to obtaining expert referees' comments on the article. The problem is compounded in papers from third-world countries where bio-statisticians may not be available in all institutions to guide clinicians as to the selection of proper techniques.Methods: The present paper deals with the various techniques of survival analysis and their interpretation, using a modal data set of malignant upper-aerodigestive tract melanoma patients treated in the Regional Cancer Centre, Trivandrum since 1982.Results: The Kaplan-Meier method was found to be the most suitable for survival analysis. The median survival time is a better method of summarizing data than the mean. Rothman's method of estimation of the confidence limit is better than Peto's method as the confidence limit for survival probability tends to go beyond the range of 0-1.0 when calculated by Peto's method, especially when the sample size is small.Conclusion: The results from the present study suggest that survival analysis should be carried out by the Kaplan-Meier method. The median survival time should be provided wherever possible, rather than relying on mean survival. Confidence limits should be calculated as a measure of variability. A suitable rank test should be used to compare two or more survival curves, rather than a Z-test. Stratified analysis and Cox's model, when stratified analysis fails, can be used to define the impact of prognostic factors on survival.