Practical exercise 6


The bone marrow transplantation data are described in example 1.13, and in practical exercise 5 it is described how you can read the data into R.


In practical exercise 5 we used a Markov illness-death model analyze the bone marrow transplantation data. In this exercise we will look at relapse-free survival (i.e. survival without relapse) without considering whether the platelet count of a patient returns to a self-sustaining level or not. Thus we only consider the censored survival time coded as T2 in the data file with corresponding censoring indicator DF.


The bone marrow transplantation patients may be classified into the three disease groups:

  • acute lymphoblastic leukemia (ALL)
  • low-risk acute myelotic leukemia (AML low-risk)
  • high risk acute myelotic leukemia (AML high-risk)

The disease group is coded as g in the data file.


a) Make Nelson-Aalen plots for the cumulative relapse-free hazard rates for the three disease groups.


b) Test the null hypothesis that the relapse-free hazard rates are the same for the three groups.