We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ME

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Methodology

Title: Bayesian inference on the number of recurrent events: A joint model of recurrence and survival

Abstract: The number of recurrent events before a terminating event is often of interest. For instance, death terminates an individual's process of rehospitalizations and the number of rehospitalizations is an important indicator of economic cost. We propose a model in which the number of recurrences before termination is a random variable of interest, enabling inference and prediction on it. Then, conditionally on this number, we specify a joint distribution for recurrence and survival. This novel conditional approach induces dependence between recurrence and survival, which is often present, for instance due to frailty that affects both. Additional dependence between recurrence and survival is introduced by the specification of a joint distribution on their respective frailty terms. Moreover, through the introduction of an autoregressive model, our approach is able to capture the temporal dependence in the recurrent events trajectory. A non-parametric random effects distribution for the frailty terms accommodates population heterogeneity and allows for data-driven clustering of the subjects. A tailored Gibbs sampler involving reversible jump and slice sampling steps implements posterior inference. We illustrate our model on colorectal cancer data, compare its performance with existing approaches and provide appropriate inference on the number of recurrent events.
Comments: 56 pages, 11 figures
Subjects: Methodology (stat.ME)
Journal reference: Statistical Methods in Medical Research 31 (2022) 139-153
DOI: 10.1177/09622802211048059
Cite as: arXiv:2005.06819 [stat.ME]
  (or arXiv:2005.06819v2 [stat.ME] for this version)

Submission history

From: Willem van den Boom [view email]
[v1] Thu, 14 May 2020 08:59:01 GMT (396kb,D)
[v2] Tue, 3 Aug 2021 09:55:38 GMT (1293kb,D)

Link back to: arXiv, form interface, contact.