Competing risks
A comprehensive introduction to competing-risks models for predicting absolute risks of disease and prognosis from time-to-event data.
What this course covers
Register interest
This course provides a comprehensive introduction to competing-risks models for predicting absolute risks of disease and prognosis using time-to-event data. Participants gain a practical understanding of the key concepts, estimation methods, and software implementation across three sessions spread over three days.
By the end, participants understand when competing risks must be considered, the independence assumption and its impact on estimation and interpretation, and the cause-specific cumulative incidence function and its estimation. The primary software is Stata (version 17+), with full R solutions provided.
Full course title: Competing risks: An introduction to concepts, methods & software
Try it in your browser: the multi-state & competing-risks explorer — live cumulative-incidence curves and state occupancy.
At a glance
What you'll learn
Session 1 · Foundations
Survival analysis recap, the competing-risks setting, the independence assumption, all-cause vs cause-specific mortality, and the cause-specific cumulative incidence function.
Session 2 · Estimation
The Aalen-Johansen estimator, flexible parametric survival models, and prediction of the cause-specific CIF.
Session 3 · Extensions
Non-proportional hazards, Fine and Gray models, weighted estimates, and a summary of the key concepts.
Audience & prerequisites
Who it's for
Epidemiologists, statisticians, physicians, public health specialists or anyone with an interest in methods for competing risks.
Prerequisites
Basic knowledge of the fundamentals of epidemiology and biostatistics, and comfort fitting and interpreting statistical models in epidemiology (e.g. linear, logistic, Poisson, or Cox regression). Prior basic knowledge of Stata is assumed.