Joint frailty models for zero-inflated recurrent events in the presence of a terminal event

• Published in: Biometrics Vol. 72; no. 1; pp. 204 - 214
• Main Authors: Liu, Lei, Huang, Xuelin, Yaroshinsky, Alex, Cormier, Janice N.
• Format: Journal Article
• Language: English
• Published: United States Blackwell Publishing Ltd 01.03.2016; International Biometric Society
• Subjects: Acquired immune deficiency syndrome; AIDS; BIOMETRIC PRACTICE; Biometrics; biometry; Cancer; Clinical trials; Computer Simulation; Cure model; Data Interpretation, Statistical; death; Finite Mixture; Hazards; Humans; Latent Class; logit analysis; Longitudinal Studies; Models, Statistical; Neoplasm Recurrence, Local - mortality; patients; Prevalence; Reproducibility of Results; Risk Assessment - methods; Sarcoma; Sarcoma - mortality; Sensitivity and Specificity; Statistical models; Survival Analysis; Tumors; Survival analysis; Latent Class; Cure model; Finite Mixture
• ISSN: 0006-341X; 1541-0420; 1541-0420
• DOI: 10.1111/biom.12376

Recurrent event data arise frequently in longitudinal medical studies. In many situations, there are a large portion of subjects without any recurrent events, manifesting the "zero-inflated" nature of the data. Some of the zero events may be "structural zeros" as patients are unsusceptible to recurrent events, while others are "random zeros" due to censoring before any recurrent events. On the other hand, there often exists a terminal event which may be correlated with the recurrent events. In this article, we propose two joint frailty models for zero-inflated recurrent events in the presence of a terminal event, combining a logistic model for "structural zero" status (Yes/No) and a joint frailty proportional hazards model for recurrent and terminal event times. The models can be fitted conveniently in SAS Proc NLMIXED. We apply the methods to model recurrent opportunistic diseases in the presence of death in an AIDS study, and tumor recurrences and a terminal event in a sarcoma study.