Subgroup analysis methods for time-to-event outcomes in heterogeneous randomized controlled trials