论文部分内容阅读
Competing risks framework,where an individual fails due to multiple causes,is frequently available in biomedical studies.Recently,joint modeling of longitudinal measurements and survival endpoints of competing risks has become prominent.We proposed a joint model considering fullyspecified subdistribution model(Ge and Chen,2012)and longitudinal measurements.The proposed joint model includes a linear mixed effect submodel and a fully-specified subdistribution submodel linked through a shared random effect.In addition,we introduced an observed binary variable for individuals with unobservable and missing at random longitudinal measurements to include the competing risks information with no longitudinal measurement.