摘要
目的探讨联合脆弱Copula模型在肺癌患者反复住院、死亡的影响因素分析。方法通过收集确诊的非小细胞肺癌患者反复住院的随访资料,构建联合脆弱Copula模型,采用极大惩罚对数似然函数进行参数估计。结果联合脆弱Copula模型利用Copula函数描述了由于未测量的复发水平上的协变量所导致的残差相依性,利用脆弱项引入了个体水平上的复发事件与终止事件间及复发事件间的相关性,客观地评价了肺癌患者疾病进展及死亡的影响因素,结果解释合理,软件实现方便。结论联合脆弱Copula模型能够深入地分析和解释肺癌患者疾病进展及结局的随防资料所蕴含的信息,进一步拓展联合脆弱模型在含终点的临床复发事件数据研究中的应用,为医学实践中肿瘤患者疾病进展的随访研究提供方法学支持。
Objective Investigating influence factors of tumour progression and death of lung cancer using joint frailty-copula models and R software.Methods Collecting the follow-up data of non-small cell lung cancer,Constructing the joint frailty-copula model,performing the maximum penalized likelihood to estimate the parameters,and analyzing the data using the R software joint.Cox3.4 package.Results The joint frailty-copula model describes residual dependence induced by unmeasured recurrence-specific covariats using the Copula function,the model introduces patient-level dependence between recurrence and death,and dependence among recurrences using the frailty.Objectively evaluated the influence factors of disease progression and death of lung cancer patients,the results are reasonable interpretation,software is convenient and also available.Conclusion The joint frailty-copula model can deeply analyze and explain the information of lung cancer patients’s progression and death,and further expand the application of the model.The model can analyze data having information on terminal event time.Provide the methodological support for the similar follow-up study of disease progression in tumor patients in medical practice.
作者
罗天娥
李淼
郭强
于智凯
赵晋芳
段燕
Luo Tiane;Li Miao;Guo Qiang(Department of Health Statistics,Shanxi Medical University(030001),Taiyuan)
出处
《中国卫生统计》
CSCD
北大核心
2020年第5期654-658,共5页
Chinese Journal of Health Statistics
基金
山西省面上自然基金项目(201801D121210)
国家青年科学基金项目(81001294)。
