
Treating non-proportional hazards as a yes/no problem risks obscuring clinically relevant time-varying effects. We thus argue for approaches and interpretations grounded in temporal dynamics and clinically meaningful estimands. doi.org/10.1016/j.jhep…












