Association between Multiple Recurrent Events with Multivariate Modeling: A Retrospective Cohort Study

Area of Science:

• Biostatistics
• Oncology
• Epidemiology

Background:

• Recurrent event data are common in biomedical research, such as infections or hospitalizations.
• Studies often involve multiple types of events, necessitating advanced statistical approaches.
• Multivariate joint modeling can identify associations between different event types.

Purpose of the Study:

• To identify the association between two types of recurrent events in breast cancer patients.
• To apply multivariate joint modeling to clinical data for analyzing recurrent events.
• To investigate the relationship between tumor grade and recurrence risk.

Main Methods:

• Retrospective cohort study of 342 breast cancer patients (2006-2015).
• Utilized a joint frailty model implemented in the R software's Frailty package.
• Patients were monitored for at least six months post-diagnosis.

Main Results:

• Three-year and five-year survival rates were 0.79 and 0.68, respectively, when death was the terminal event.
• The joint frailty model indicated an increased risk of local and metastatic recurrences for patients with tumor grades above I.
• Significant variance in the frailty component for metastases suggests individual variability in risk.

Conclusions:

• Patients with similar predictive variables may face varying levels of metastases risk.
• The significant frailty variance for metastases implies inherent individual differences in recurrence susceptibility.
• Caution is advised when interpreting results due to the low frequency of observed recurrence types.