Hi,

I am analysing a small trial where the primary outcome is time to event (recovery). All individuals are followed up until either the event occurs (recov==1), or the individual dies (recov==0; no individuals are censored), so I think the best thing to do is to produce the CIF in the presence of competing risks rather than using Kaplan Meier methods.

I am trying two methods to produce the CIF curves & am having trouble interpreting them or choosing which is best to use. I provide the data below using dataex.

I have tried the following using the Fine & Grey model:

Code:
stset days, failure(recov==1)
stcrreg i.group, compete(recov==0)
stcurve, cif at1(group=0) at2(group=1)
Apologies if this is obvious, but I'm not sure why the curves don't end at the actual proportion experiencing the event in each; group (tab recov group, col - 72% and 94% in this example).

Secondly, I have tried using the following method using stcompet (ssc install stcompet) as described here:

Code:
stset days, failure(recov==1)
stcompet cif = ci, compet1(0) by(group)
gen cif_local_drug0 = cif if group==0
gen cif_local_drug1 = cif if group==1
twoway line cif_local_drug1 cif_local_drug0 _t, connect(step step) sort
I find this intuitive as the proportions are equal to the total proportions of events after the follow-up is complete. However, I get strange dips in the CIF which I don't understand.

Does anyone have any advice on how I can proceed? Thank you very much in advance.

Megan



Code:
* Example generated by -dataex-. To install: ssc install dataex
clear
input float(group recov) byte days
0 0  2
0 0  2
0 0  6
0 0  8
0 1  9
0 1 10
0 1 12
0 1 14
0 1 16
0 1 16
0 0 16
0 1 18
0 1 18
0 1 18
0 1 18
0 1 18
0 1 20
0 1 20
0 0 20
0 1 22
0 0 24
0 1 30
0 1 32
0 1 36
0 1 38
1 1  6
1 1  8
1 1  8
1 1  8
1 1  8
1 1  8
1 1  8
1 0 10
1 1 10
1 1 10
1 1 10
1 1 10
1 1 10
1 1 11
1 1 11
1 1 12
1 1 12
1 1 12
1 1 12
1 1 12
1 0 12
1 1 14
1 1 14
1 1 14
1 1 14
1 1 16
1 1 18
1 1 18
1 1 20
1 1 22
1 1 22
1 1 28
end
label values group group2
label def group2 0 "control", modify
label def group2 1 "treat", modify