Chapter 14 Survival Analysis

These notes rely on the Survival Analysis in R DataCamp course, STHDA, and Applied Survival Analysis Using R (Moore 2016).

Survival analysis models time to event. Whereas linear regression outcomes are assumed to have a normal distribution, time-to-event outcomes have a Weibull or unknown distribution. Survival analysis models also deal with censoring (unknown starting event and/or ending event). These factors make survival analysis more complicated than linear regression.

Most survival analyses use the survival package for modeling and the survminer package for visualization.

library(tidyverse)
library(survival)
library(survminer)

A typical survival analysis uses Kaplan-Meier plots to visualize survival curves, log-rank tests to compare survival curves among groups, and Cox proportional hazards regression to describe the effect of variables on survival.

References

Moore, Dirk F. 2016. Applied Survival Analysis Using R. 1st ed. New York, NY: Springer. https://eohsi.rutgers.edu/eohsi-directory/name/dirk-moore/.