Applied Survival Analysis in R: Kaplan–Meier Curves, Cox Models & Competing Risks

Aim

Workshop Objectives

Workshop Structure

Day 1: Survival Data, Censoring & Kaplan–Meier Curves Using R
Introduction to time-to-event data using clinical and cohort examples.
Understanding event times, right-censoring and how censoring is represented in survival data.
Interpreting survival probability and the Kaplan–Meier estimator.
Introduction to the colon trial dataset and its treatment arms.
Reshaping trial data to one row per patient for analysis.
Reading survival curves and interpreting median survival.
Introduction to the log-rank test for comparing treatment groups.
Hands-On 1: Kaplan–Meier Analysis and Treatment-Group Comparison Using R
Work through a small example by hand, compare it with R output, plot survival curves for the colon trial arms, estimate median survival and run a log-rank test.
Tools: R, Google Colab with an R runtime or local R/RStudio, survival package.
Output: Kaplan–Meier plot, median survival estimates and a brief interpretation of the treatment-group comparison.

Day 2: Cox Regression, Confounding & Model Assumptions Using R
Review of survival curves, censoring and the log-rank test.
Understanding hazard as an instantaneous event rate and distinguishing it from survival probability.
Interpreting hazard ratios, confidence intervals and reference groups.
Recognizing potential confounding using age and lymph-node involvement as examples.
Comparing crude and adjusted Cox proportional-hazards models.
Interpreting a forest plot of model estimates.
Understanding the proportional-hazards assumption and why it should be checked.
Hands-On 2: Crude and Adjusted Cox Models Using R
Compare treatment arms by age and node involvement, calculate crude event rates, fit simple and adjusted Cox models, create a forest plot and check the proportional-hazards assumption.
Tools: R, Google Colab with an R runtime or local R/RStudio, survival package, ggplot2.
Output: Cox model results with hazard ratios and confidence intervals, forest plot and proportional-hazards diagnostic.

Day 3: Competing Risks, Cumulative Incidence & Reporting
Understanding competing events, such as death before recurrence can occur.
Recognizing when standard Kaplan–Meier methods may overestimate the probability of a specific event.
Interpreting cumulative incidence as the probability of an event in the presence of competing events.
Counting competing events and comparing naive estimates with competing-risk estimates.
Examining recurrence risk at selected time points and in an older-patient subgroup.
Checking that event probabilities are interpreted consistently.
Reviewing clear reporting of sample size, median survival, hazard ratios, confidence intervals, model checks and competing events.
Hands-On 3: Competing-Risk Estimates and Results Reporting Using R
Count competing events in the colon trial, compare standard and competing-risk recurrence estimates at 1, 3 and 5 years, repeat the comparison for participants aged 70 and over, and prepare a short results summary.
Tools: R, Google Colab with an R runtime or local R/RStudio, survival package, cmprsk package.
Output: Competing-event summary, cumulative-incidence comparison and a concise analysis report with key estimates and model limitations.

Who Should Enrol?

Important Dates

Workshop Outcomes

Fee Structure