CAJ | 학술논문

目的以实例阐述生存分析分段指数模型的拟合及SAS实现.方法将生存时间划分为几个区间,利用SAS中的PROC GENMOD过程或PROC LIFEREG过程,对生存资料进行分段指数模型拟合,并通过参数估计结果,计算各时间区间的风险率及生存率.结果该资料拟合分段指数分布的结果比较满意,三个时间区间的死亡风险均不相同,且第一个和第二、三个区间的差别有统计学意义(P<0.05).结论分段指数模型不仅模型形式简单,参数易于估计,若拟合得当可提高统计效率.
목적 이실례천술생존분석분단지수모형적의합급SAS실현.방법 장생존시간화분위궤개구간,이용SAS중적PROC GENMOD과정혹PROC LIFEREG과정,대생존자료진행분단지수모형의합,병통과삼수고계결과,계산각시간구간적풍험솔급생존솔.결과 해자료의합분단지수분포적결과비교만의,삼개시간구간적사망풍험균불상동,차제일개화제이、삼개구간적차별유통계학의의(P<0.05).결론 분단지수모형불부모형형식간단,삼수역우고계,약의합득당가제고통계효솔.
Objective To introduce the fitfng of piecewise exponential model in survival analysis and its corresponding SAS proce-dures. Methods Divide the survival time scales into several intervals first. Then by PROC GENMOD procedure or PROC LIFEREG procedure in SAS, piecewise exponential model can be built for the survival data. Hazard function and survival function of each time interval were calculated based on the results of parameter estimation. Results The result of fitting was relative satisfactory. The hazards of death in three intervals were different. Differences between the first and the second,the first and the third interval were statistically significant. Conclusion Not only is the form of the piecewise exponential model simpler,but also it's easier to estimate param-eters. If the model is appropriately fitted,the statistical efficiency can be im-proved.