SemiCompRisks

SemiCompRisks implements analysis of semi-competing risks data using illness-death multi-state models to quantify relationships between non-terminal and terminal events in medical and epidemiological studies.

Key Features:

• Model Flexibility: Supports independent and cluster-correlated semi-competing risks using an illness-death multi-state model framework.
• Model Specifications: Supports accelerated failure time and proportional hazards regression, parametric or non-parametric baseline survival functions, parametric or non-parametric random-effects distributions for clustered data, and Markov or semi-Markov specifications for the terminal event following a non-terminal event.
• Estimation Methods: Implements Bayesian estimation for general models and maximum likelihood estimation (MLE) for selected parametric models.
• Complementary Analyses: Includes functions for univariate survival analysis.

Scientific Applications:

• Medical research: Enables analysis of disease progression and survival where non-terminal events (e.g., progression) and terminal events (e.g., death) interact.
• Epidemiology: Facilitates population-level inference about the relationship between non-terminal events and mortality in cohort and longitudinal studies.

Methodology:

Uses the illness-death multi-state model with options for accelerated failure time or proportional hazards regression, parametric or non-parametric baseline survival, parametric or non-parametric random effects for clustered data, Markov or semi-Markov transition specifications, and estimation by Bayesian methods or MLE for selected parametric cases.