论文部分内容阅读
[目的]探讨如何利用最小P值法研究变量的最佳界值,为临床预后因素的分析提供方法学参考。[方法]以76例接受铂类化疗的胃癌患者的生存时间和ERCC1mRNA的表达水平为例,经过多次对数秩检验得到最大χ2值及ERCC1的界值,并利用近似校正法和Bootstrap法对相应的P值进行校正。[结果]最大χ2值为32.441,对应的ER-CC1界值为0.4632。两种校正法得到的P值均小于0.001,说明ERCC1表达水平与胃癌含铂方案化疗患者生存时间有相关性。[结论]ERCC1mRNA的表达水平可以作为以铂类为基础药物化疗的胃癌患者的生存时间的预测分子,最小P值法适用于肿瘤的临床预后因素研究。
[Objective] To explore how to study the best cut-off value of variables by using the minimum P-value method, and provide a methodological reference for the analysis of clinical prognostic factors. [Method] The survival time and the expression level of ERCC1 mRNA in 76 gastric cancer patients receiving platinum-based chemotherapy were taken as examples. After multiple log rank tests, the maximum χ2 value and the cut-off value of ERCC1 were obtained. Approximate correction method and Bootstrap method The corresponding P value is corrected. [Results] The maximum χ2 value was 32.441 and the corresponding ER-CC1 cutoff value was 0.4632. The P value of the two calibration methods were less than 0.001, indicating that ERCC1 expression levels and platinum-containing chemotherapy in patients with gastric cancer survival time has a correlation. [Conclusion] The expression level of ERCC1 mRNA can be used as a prognostic factor for the survival time of patients with gastric cancer based on platinum-based chemotherapy. The minimum P value is suitable for clinical prognostic factors of tumors.