Hello.

I am currently working on a study of depression in which the goal is to evaluate onset age curves by gender (cumulative incidence). Onset ages for depression were established by four separate diagnostic interviews occurring between the ages of 14 and 33. Briefly; 1709 adolescents were interviewed at T1, 1507 of the 1709 (88%) were interviewed at T2, 941 of the 1507 (63%) were interviewed at T3, and 816 of the 941 (86%) were interviewed at T4.

Eligible participants for T3 included all persons with a positive history of a psychiatric diagnosis at T2 (n = 644) and a randomly selected subset of participants with no history of mental disorder at T2 (n = 457 of 863 persons, which also included all non-white T2 participants to enrich the ethnic representativeness and diversity of the sample). Of these 1,101 eligible persons, 941 completed the T3 diagnostic interview.

In other analyses based on the 816 T4 participants, a weighting procedure was used whereby Caucasian participants without a lifetime diagnosis by T2 were assigned a weight of 2.05, a value that reflects the probability of this subgroup being sampled at T3. For the current analysis, however, we wish to include all available data (not limit ourselves to the T4 sample).

Considering the nature of the data set and the combination of censoring and stratified sampling procedures at T3, are there any suggestions as to what software/procedures be used to estimate survival curves?

Thanks in advance for your insights!

-d