# Shared Frailty Assumptions

Do we need to check the proportional hazards assumption when running a shared frailty model for clustered data?