Need help with follow up analysis

[Kashvig]

Hello, I am trying to do some survival analysis and have a dataset that looks like this:

As  you can see the first 6 lines are from the same person. fu_6m to fu_5 are the follow up status from 6 months to 5 years. 0 is dead and 1 is alive. 

I need help with having one unique observeration per person and having a variable for follow up status at 6 month, 1 year, 2 year .... 5 year

[PaigeMiller]

Since you don't have unique records for each person, you have six different records, what do you want to see as the output for each person?

[Kashvig]

I would like to combine the six records and see for each person something like this:

pers_id      px_id     fu_6m          fu_1         fu_2          fu_3          fu_4       fu_5

123            123        1                   1               1               0              0             0

124             124        1                   1               1               1              0             0

Where 123 has died at follow up at 3 years and 124 has died at follow up at 4 years

[PaigeMiller]

---

@Kashvig wrote:

I would like to combine the six records and see for each person something like this:

pers_id      px_id     fu_6m          fu_1         fu_2          fu_3          fu_4       fu_5

123            123        1                   1               1               0              0             0

124             124        1                   1               1               1              0             0

Where 123 has died at follow up at 3 years and 124 has died at follow up at 4 years

---

I would like you to show the desired output from the input data set, not the desired output from two individuals who were not in the input data set.

[Kashvig]

If I am understanding your Q correctly I would want the output data for lines 1-6 look like this

pers_id    px_id.        fu_6m.     fu_1.   fu_2.     fu_3.    fu_4.    fu_5

4736884   715723       1              1          1         1            1          1

[PaigeMiller]

So are you asking for the sum of the six rows for each patient?

[Kashvig]

I think so, yes

[PaigeMiller · Accepted solution]

proc summary data=have nway;
    class pers_id px_id;
    var fu_6m--fu_5;
    output out=want sum=;
run;

[Kashvig]

Thank you this worked!

[StatDave]

See the example titled "Comparison of the Marginal and Random Effect Models for Binary Data" in the PROC GEE documentation in the SAS/STAT User's Guide. This shows how the data can be rearranged so that there is observation per measurement. The repeated binary measures data can then be modeled with either a Generalized Estimating Equations model or a random effects model.