Pairwise comparisons and adjustment for multiple comparisons in stratified test of equality of survivor functions

[Meghan Bezerra]

Hello,

I'm using Stata 18 with the following data:

* Example generated by -dataex-. For more info, type help dataex
clear
input int id byte divorce float(educ_hus Black_hus race_diff) double(marriage_dur age) float age_cat double income byte income_quartiles
  9 0 1 0 0  10.546   26.6365903358909 1   75 3
 11 0 0 0 0  34.943  26.33688587358758 1   59 2
 13 1 0 0 0   2.834 23.171812946223962 0 53.5 1
 15 1 0 0 0  17.532 24.960095368077482 0 64.5 2
 33 0 1 0 0   1.418  26.17530695456667 1   60 2
 36 0 0 0 0  48.033  26.86812459610563 1   64 2
 43 0 2 0 0  16.706 28.850875999733052 1   89 4
 47 0 0 0 0  24.999  22.02641852754293 0   39 1
 50 0 0 0 0  24.999  22.78547072959178 0   78 4
 56 0 0 1 0   3.869 24.965775644983125 0   63 2
 63 0 1 1 0   7.732  29.52430847817645 1   36 1
 66 1 1 0 0  5.2105 27.601385355995284 1 95.5 4
 70 0 1 0 0  15.444  28.20856646555208 1   51 1
 77 0 2 0 0   4.085  26.69471565374245 1   65 2
 80 0 0 0 0  17.333 28.967617641364775 1   47 1
 84 0 1 0 0  16.331 23.765405394717792 0  108 4
 87 0 1 0 0  35.335 27.591393217926697 1   89 4
 90 0 0 0 0   37.67 25.949986446108696 1   58 2
 91 1 1 0 0 19.1865  24.79527934587675 0 90.5 4
 94 0 1 0 0  22.697 29.284290621090683 1   76 3
109 0 0 0 0  50.776   29.6396336137035 1   74 3
111 0 0 0 0  37.495  24.26853923650965 0   49 1
114 0 0 0 0  30.738  23.02254705185942 0   72 3
116 0 0 0 0  48.654  25.51933961395623 1   56 2
118 0 0 0 0  33.024 25.570099081052152 1   65 2
120 0 0 0 0  36.668 27.246295948390504 1   74 3
122 0 2 0 0  27.992 26.555117034315536 1   72 3
124 0 2 0 0  27.789 25.259106173207545 0   83 4
128 1 1 0 0  3.2525 23.802624487535013 0 73.5 3
129 1 1 0 0    5.67 25.277288385082883 0 89.5 4
end
label values divorce Black
label values Black_hus Black
label values race_diff Black
label def Black 0 "No", modify
label def Black 1 "Yes", modify
label values educ_hus educ_hus
label def educ_hus 0 "<12 years", modify
label def educ_hus 1 "12-15 years", modify
label def educ_hus 2 "16+ years", modify
label values age_cat age
label def age 0 "25 and under", modify
label def age 1 "Over 25", modify

I have stset my data like so:

stset marriage_dur, failure(divorce) id(id)

and am exploring differences in survival functions stratified by categorical variables in the dataset. I'm having no trouble using the log-rank test with binary categorical variables like age_cat and race_diff, but I cannot find a straightforward way to test the equality of survivor functions stratified by educ_hus and income_quartiles with output for pairwise comparisons and adjustment for multiple comparisons. If anyone can share with me how to approach this, I would appreciate it.

-Meghan