---

Abstract
Survival outcomes in clinical trials are routinely summarized using hazard ratios, largely due to the widespread adoption of the Cox proportional hazards model. Despite their popularity, hazard ratios can be difficult to interpret clinically and rely on assumptions—most notably proportional hazards—that are often violated in practice. In recent years, alternative summary measures have gained increasing attention, including approaches based on restricted mean survival time (RMST), which may offer improved interpretability and efficiency in certain settings. This seminar will discuss the dominance of hazard ratios in trial reporting, outline some of their limitations, and introduce alternative estimands for summarizing treatment effects in survival analysis. Using examples from clinical trial research, I will illustrate situations in which alternative summary measures may provide a more meaningful, powerful and/or robust assessment of treatment benefit.

Il seminario si terrà presso l'Aula F del Palazzo delle Scienze (Via Ospedale, 72, Cagliari) e potrà essere seguito anche su Microsoft Teams al link:

STATISTICS SEMINARS - 25 Mag @UniCa (DMI) - Speaker: Matteo Quartagno (UCL) | Partecipazione alla riunione | Microsoft Teams