I have a right-censored data set. I plot a Kaplan–Meier survival curve using R: Kaplan-Meier survival curve

The y-values are easy to interpret, e.g. about 45% of the population will survive 500 days.

Now I plot the Nelson–Aalen cumulative hazard curve by setting the parameter fun="cumhaz" in the plot function: Nelson–Aalen cumulative hazard curve Do the y-values have any intuitive interpretation in this case? Are there any advantages to the latter plot in a presentation?

  • 1
    $\begingroup$ One "intuitive" interpretation of the cumulative hazards I know of is: if we assume that a subject under observations is immediately revived when he dies then the cumulative hazards at time t measures the expected amount of observed deaths. $\endgroup$ May 30, 2017 at 7:51
  • $\begingroup$ Does this mean that when y=1, x is the expected (mean) lifetime? I thought it was not possible to calculate the expected lifetime when censored data is present. $\endgroup$
    – matthiash
    May 31, 2017 at 10:30
  • $\begingroup$ No that doesn't seem right to me, but I dug around a bit and found this stackexchange post which explains the cummulative hazards better than I can, stats.stackexchange.com/questions/60238/… $\endgroup$ May 31, 2017 at 13:14


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Browse other questions tagged or ask your own question.