I have 5 different kinds of repeated transitions nested under individual dyads. For any particular transition, the 5 different types of transitions are competing risks, only 1 will happen, which will censor the other 4. I have run this model in Mplus using Cox regression using basehazard = off and allowing 4 of the 5 types of transitions to differ from the baseline hazard by a constant. I am very interested in the dyad level correlations between the different types of transitions. I would like to make this model more general by allowing each transition type to have its own baseline hazard. Can Mplus allow different baseline hazards when the transitions are nested under dyads?