Reflection on modern methods: competing risks versus multi-state models

TL;DR

This paper compares competing risks and multi-state models in survival analysis, emphasizing their respective strengths and applications, especially in complex scenarios with non-terminal events, illustrated through a Bayesian recurrent hip fracture study.

Contribution

It clarifies the relationship between competing risks and multi-state models and demonstrates the advantages of multi-state models for complex epidemiological data analysis.

Findings

Multi-state models handle non-terminal, sequential events effectively.

Bayesian methods enhance inference in recurrent event studies.

Multi-state models provide richer insights in complex survival scenarios.

Abstract

Survival competing risks models are very useful for studying the incidence of diseases whose occurrence competes with other possible diseases or health conditions. These models perform properly when working with terminal events, such as death, that imply the conclusion of the corresponding study. But they do not allow the treatment of scenarios with non-terminal competing events that may occur sequentially. Multi-state models are complex survival models. They focus on pathways defined by the temporal and sequential occurrence of numerous events of interest and thus they are suitable for connecting competing non-terminal events as well as to manage other survival scenarios with higher complexity. We discuss competing risks within the framework of multi-state models and clarify the usefulness of both models for analysing epidemiological data. We highlight the power of multi-state models through a real-world study of recurrent hip fracture from Bayesian inferential methodology.