Survival analysis using R
>eR-BioStat
Topics
The slides in this page are organised in 6 chapters that cover all the topics in the course ``Survival analysis using R: an introduction'' or “Basic Skills in Bootstrap using R”. All examples are illustrated using the R software.
Make sure that the following R packages are installed:
-
library(survival)
-
library(eventtimedata)
Useful R functions include:
-
survfit()
-
survdiff()
-
coxph()
The course in slides format
Introduction and background
This chapter covers the following topics:
-
Basic definitions of survival distributions and hazard functions
-
Types of censoring
-
Parametric survival distributions
Proportional hazards regression: special topics
-
Graphical diagnostics for the Cox model
-
Regression with correlated event-time data
Non-parametric estimation of a survival distribution
This chapter covers the following topics
-
The Kaplan-Meier estimator
-
The cumulative hazard estimator
-
Estimating standard errors, including Greenwood's formula and the delta method for transformations
-
Confidence intervals and confidence bands for survival distributions
-
Alternatives to median survival: restricted mean survival
Designing a survival study
This chapter cover the following topics:
-
Power and sample size calculations for survival distributions and proportions in two groups
-
Adjusting for staggered arrival and loss to follow-up
-
Software in R for trial design
-
Fixed sample designs vs sequential designs
Significance tests with censored data
This chapter covers the following topics
-
The log-rank test for two samples
-
The Tarone-Ware family of weighted log-rank tests for non-proportional hazards
-
Tests for more than two groups
-
Stratified tests
Proportional hazards regression: basics
This chapter covers the following topics:
-
The Cox proportional hazards model
-
Partial likelihood estimation and inference
-
The link between the Cox model and the log-rank test
-
Time-varying covariates