gsem for joint modelling with binary longitudinal outcome

[Gianfranco Di Gennaro]

Dear all, I have a wide shaped (thanks to Clyde Schechter again!) dataset in which I collected the time-to event data of the drop out of 5 drugs, in different hospitals and with covariates (adverse events, therapeutci success and so on.)

Simplifying my variables are:

• 5 different drugs (variable: "drug")
• Adverse events (time dependent dichotomous variable: "AdEvents")
• Therapeutic success (time dependent dichotomous variable: "dicOutcome")
• drug dropout: "intherapy == 0" (absence of outcome)
• followup time: "followup"
• variable order of followup (first, second, third ... followup: variable "t")
• center (variable "hospital")

The aims of the study are:

• to estimate dropout time in the 5 drugs, accounting for the different hospitals)
• to estimate longitudinalli adverse events (variable "AdEvents") and therapeutci succes (variable "dicOutcome")

I ran a shared frailty model that worked very well.
Anyway, I'd like to investigate dropout, adverse events and therapeutc success simultaneously.

My idea is to run two joint models:

• dropout (time-to-event) and adverse event (longitudinal variable) and accounting for Sex, Age and therapeutc effect
• dropout (time-to-event) and therapeutc effect (longitudinal variable) and accounting for Sex, Age and adverse event

I did it by the command "stjm". The problem is that stjm cannot accomodate binary longitudinal variables.

The command to use, as I can see, is "gsem"

I'm having a lot of troubles in specifying it. Does anybody know what the correct code is?

I declared time-to-event data:
stset followup, id(id) failure(intherapy == 0).

The I tried many times but it never worked.

Thanks in advance.
I hope this topic will be useful not only to me

* Example generated by -dataex-. For more info, type help dataex
clear
input float(id t followup intherapy) int age byte(sex drug) float dicOutcome byte AdEvents str5 hospital
 1 1         6 1 39 0 5 0 0 "IS_"
 1 2        12 1 39 0 5 0 0 "IS_"
 1 3 12.866667 1 39 0 5 0 0 "IS_"
 1 4        36 1 39 0 5 0 0 "IS_"
 2 1         6 1 28 1 1 0 0 "IS_"
 2 2        12 1 28 1 1 1 0 "IS_"
 2 3 20.766666 1 28 1 1 1 0 "IS_"
 2 4        36 1 28 1 1 1 0 "IS_"
 3 1         6 1 31 1 1 0 0 "IS_"
 3 2        12 1 31 1 1 0 0 "IS_"
 3 3      21.3 1 31 1 1 0 0 "IS_"
 3 4        36 1 31 1 1 0 0 "IS_"
 4 1         6 1 31 1 3 0 0 "IS_"
 4 2        12 1 31 1 3 1 0 "IS_"
 4 3      14.2 1 31 1 3 1 0 "IS_"
 4 4        36 1 31 1 3 1 0 "IS_"
 5 1         6 1 21 0 1 0 0 "IS_"
 5 2        12 1 21 0 1 0 0 "IS_"
 5 3        24 1 21 0 1 0 0 "IS_"
 5 4 33.466667 1 21 0 1 0 0 "IS_"
 6 1         6 1 28 1 1 0 0 "IS_"
 6 2        12 1 28 1 1 0 0 "IS_"
 6 3        20 0 28 1 1 0 0 "IS_"
 7 1         6 1 27 1 1 0 0 "IS_"
 7 2        12 1 27 1 1 1 0 "IS_"
 7 3        24 1 27 1 1 1 0 "IS_"
 7 4 33.466667 1 27 1 1 1 0 "IS_"
 8 1         6 1 30 1 3 0 1 "IS_"
 8 2        12 1 30 1 3 0 1 "IS_"
 8 3 33.466667 1 30 1 3 0 1 "IS_"
 8 4        36 1 30 1 3 0 1 "IS_"
 9 1         6 1 31 0 1 0 0 "IS_"
 9 2        12 1 31 0 1 1 0 "IS_"
 9 3        24 1 31 0 1 1 0 "IS_"
 9 4 33.466667 1 31 0 1 1 0 "IS_"
10 1         6 1 34 1 2 1 1 "IS_"
10 2        12 1 34 1 2 1 1 "IS_"
10 3        24 1 34 1 2 1 1 "IS_"
10 4      44.6 1 34 1 2 1 1 "IS_"
11 1         6 1 33 0 2 0 0 "IS_"
11 2        12 1 33 0 2 1 0 "IS_"
11 3        24 1 33 0 2 1 0 "IS_"
11 4  43.66667 1 33 0 2 1 0 "IS_"
12 1         6 1 50 0 2 1 0 "IS_"
12 2        12 1 50 0 2 1 0 "IS_"
12 3        24 1 50 0 2 1 0 "IS_"
12 4  45.63334 1 50 0 2 1 0 "IS_"
13 1         6 1 40 1 2 1 0 "IS_"
13 2        12 0 40 1 2 1 0 "IS_"
14 1         6 1 56 0 2 0 0 "IS_"
14 2       9.1 1 56 0 2 0 0 "IS_"
14 3        12 1 56 0 2 0 0 "IS_"
14 4        36 1 56 0 2 0 0 "IS_"
15 1         6 1 41 1 1 1 0 "IS_"
15 2        12 1 41 1 1 1 0 "IS_"
15 3        24 1 41 1 1 1 0 "IS_"
15 4  45.56667 1 41 1 1 1 0 "IS_"
16 1         6 1 22 1 1 0 0 "IS_"
16 2        12 1 22 1 1 1 1 "IS_"
16 3        24 1 22 1 1 0 1 "IS_"
16 4      44.6 1 22 1 1 0 1 "IS_"
17 1         6 1 58 1 1 1 1 "IS_"
17 2        12 1 58 1 1 1 0 "IS_"
17 3        24 1 58 1 1 1 0 "IS_"
17 4  45.63334 1 58 1 1 1 0 "IS_"
18 1         6 1 38 1 1 1 0 "IS_"
18 2        12 1 38 1 1 1 0 "IS_"
18 3        24 1 38 1 1 1 0 "IS_"
18 4  45.63334 1 38 1 1 1 0 "IS_"
19 1         6 1 22 0 3 1 0 "IS_"
19 2        12 1 22 0 3 1 0 "IS_"
19 3        24 1 22 0 3 1 0 "IS_"
19 4 33.466667 1 22 0 3 1 0 "IS_"
20 1         6 1 18 1 3 1 0 "IS_"
20 2        12 1 18 1 3 1 0 "IS_"
20 3        24 1 18 1 3 1 0 "IS_"
20 4  45.43333 1 18 1 3 1 0 "IS_"
21 1         6 1 20 1 3 1 0 "IS_"
21 2        12 1 20 1 3 1 0 "IS_"
21 3        24 1 20 1 3 1 0 "IS_"
21 4 33.266666 1 20 1 3 1 0 "IS_"
22 1         6 1 29 0 3 1 0 "IS_"
22 2        12 1 29 0 3 1 1 "IS_"
22 3        24 1 29 0 3 1 1 "IS_"
22 4 33.466667 1 29 0 3 1 1 "IS_"
23 1         6 1 59 1 3 1 0 "IS_"
23 2        12 1 59 1 3 1 0 "IS_"
23 3        24 1 59 1 3 1 0 "IS_"
23 4  45.63334 1 59 1 3 1 0 "IS_"
24 1         6 1 19 0 3 1 0 "IS_"
24 2        12 1 19 0 3 1 0 "IS_"
24 3        24 1 19 0 3 1 0 "IS_"
24 4 33.466667 1 19 0 3 1 0 "IS_"
25 1         6 1 43 1 3 0 0 "IS_"
25 2        12 1 43 1 3 0 0 "IS_"
25 3        13 0 43 1 3 0 0 "IS_"
26 1         6 1 40 0 3 0 1 "IS_"
26 2        12 1 40 0 3 0 1 "IS_"
26 3        13 0 40 0 3 0 1 "IS_"
27 1         6 1 28 1 3 0 0 "IS_"
end