Cox Proportional Hazards

Cox proportional-hazards regression in R β€” hazard ratios with 95% CIs via gtsummary, and the essential proportional-hazards assumption check with cox.zph().

Model survival on several predictors at once, and report hazard ratios.

Run the code

Everything in the video β€” edit it and press Run Code to run real R right here, or copy it:

Note

The first Run Code installs the packages this lesson needs (survival, gtsummary) into your browser session β€” give it 20–40 seconds (gtsummary is large β€” if it is slow in the browser, download the R script and run it locally).

Download this R script  Β·  β–Ά Open the full playground

When to use it. Use it to adjust survival for multiple covariates. A hazard ratio is only valid if the proportional-hazards assumption holds β€” always test it with cox.zph().

Read Cox proportional hazards in R.

Next: Diagnostic Test Accuracy β†’