Using the above scripts, perform a univariable Kaplan Meier analysis to determine if
ulcer influences overall survival. Hint:
survival_object ~ ulcer.
Try modifying the plot produced (see Help for ggsurvplot). For example:
- Add in a median survival line:
- Alter the plot legend:
legend.title = "Ulcer Present", legend.labs = c("No", "Yes")
- Change the y-axis to a percentage:
ylab = "Probability of survival (%)", surv.scale = "percent"
- Display follow-up up to 10 years, and change the scale to 1 year:
xlim = c(0,10), break.time.by = 1)
Create a new CPH model, but now include the variable
thickness as a variable.
- How would you interpret the output?
- Is it an independent predictor of overall survival in this model?
- Are CPH assumptions maintained?