Hey all,
Hoping for advice/clarity around cumulative incidence function plots.
I am undergoing competing risk regression analysis with 3 binary covariates and main outcome death, competing outcome relapse.
I create the cumulative incidence function plots via:
I'm wondering why the plots have 'steps' occuring for both groups (ie. for xvar1 = 0 and 1)
I understand this is a well know appearance of the CIF graphs (as seen on https://www.stata.com/features/overv...ks-regression/ and the stcurve created there) however, it has resulted in a graph where the lines are practically identical in positioning.It gives off the impression that events are occuring at same times for both groups. How is this explainable to clinicians? Is there alternative curves that people use that take into account the competing risk but have a more intuitive appearance?
Any help is much appreciated.
Hoping for advice/clarity around cumulative incidence function plots.
I am undergoing competing risk regression analysis with 3 binary covariates and main outcome death, competing outcome relapse.
I create the cumulative incidence function plots via:
Code:
stcrreg i.xvar1 i.xvar2 i.xvar3, compete(outcome=2) stcurve, cif at(xvar1=(0 1))
I understand this is a well know appearance of the CIF graphs (as seen on https://www.stata.com/features/overv...ks-regression/ and the stcurve created there) however, it has resulted in a graph where the lines are practically identical in positioning.It gives off the impression that events are occuring at same times for both groups. How is this explainable to clinicians? Is there alternative curves that people use that take into account the competing risk but have a more intuitive appearance?
Any help is much appreciated.
