repeated-measures multinomial logit vs competing risks model

[Merdock]

I need some help/advice for tackling the following problem:

The investigators I'm working with are interested in investigating the association between various covariates and the reason for termination from a kidney disease data registry. The hypothesis is that a lower proportion of ethnic/racial minorities will get kidney transplant rather than dialysis at time of end stage kidney disease onset (i.e., time of termination from registry).

Outcome: reason for termination (the dataset has this as a categorical variable with numerous categories which I narrowed down to only 4 categories: transplant, dialysis, death, other. The "other" category is the most numerous as it contains everybody terminated for any reason other than transplant/dialysis/death, plus over 300 cases who had missing value for reason for termination (these patients might still be in the registry for all we know, there's no way of knowing what happened with them), followed by transplant, dialysis and death (just a couple of cases of death).

Covariates: some are time-independent (sex, race/ethnicity) while others are measured repeatedly at 6-months visits so time-dependent (such as lab values, hypertension status, eGFR).

My initial thoughts for analysis:

A. repeated measures multinomial logistic regression analysis, given the outcome with more than 2 categories, and the time-dependent nature of some of the covariates, or

B. repeated measures competing risks/cause-specific hazards analysis

The challenges:

1. I am not sure if I even need to consider repeated measures here, as the outcome itself (reason for termination) is fixed. Nor am I sure if I even have competing risks per se (I think the idea is that we will see some racial effects where minorities get the transplant less frequently and are more often receiving dialysis but I don't really know if we need competing risks for this?)
2. If I do need to account for the repeated measures nature of some covariates, I have no experience with how to do this for multinomial or competing risks models and I haven't been able to find many resources that are easy to understand and provide specific examples for implementing in R or SAS.

Can somebody please help advise what the appropriate type of analysis would be given the study context/research hypothesis?

Thank you kindly!

[StatDave]

Seems to me that your point 1 is right. You could fit a generalized logit model in PROC LOGISTIC with the LINK=GLOGIT option to your categorical response. In the model, you could use suitable summary measures of your repeatedly measured covariates - such as averages or medians or even estimated slopes over time if that is relevant to affecting the response. With those computed summary statistics, you then would have one observation per subject and could appropriately use PROC LOGISTIC to fit the model assuming that the observations are independent.

[Merdock]

Thank you for the suggestions! It seems that this would be a two-step process with PROC LOGISTIC. I was wondering though, would it be possible to maybe use PROC GLIMMIX instead with GLOGIT link? Do you know of any good references/tutorials that might illustrate how to implement this for similar problems?

Thank you again!

[StatDave · Accepted solution]

You can use GLIMMIX to fit a logit model, but you would still need to use the GLOGIT link since your response is nominal multinomial and it will give essentially the same model as LOGISTIC.