Joint modeling approaches have been extensively employed to jointly analyze both survival and longitudinalprocesses and to investigate their relationship. Attention has mostly been paid on developing adaptive andflexible longitudinal processes based on a prespecified survival model, all chosen as multiplicative hazardsmodels and especially, very often the Cox proportional model. When the proportional hazards assumptionfails for longitudinal processes, a general class of hazards regression model maybe a useful alternative. Bycombining the Cox model and Aalen additive hazards model, we propose a joint model of semiparametricadditive-multiplicative hazards model and longitudinal processes to investigate the relationship betweensurvival time and time-dependent covariates. A pseudo joint likelihood procedure is proposed to estimate theunknown parameters through a Monte Carlo EM algorithm. A case study of evaluating the associationbetween the survival time of AIDS patients and the biomarkers, CD4 counts and viral loads demonstrates theeffectiveness of the procedure.
|Effective start/end date||1/08/17 → 31/07/18|
- Additive hazards models
- Joint modeling
- Maximum likelihood estimate
- Monte Carlo EMalgorithm
- Multiplicative hazards models
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.