A long time ago, I wondered why everyone in academia always talked about the Cox multiple logistic-regression model for proportional hazards. Now I understand.

(Professional statisticians: Okay, okay, I know the the proportional-hazards model isn't technically a multiple logistic-regression model, but that's how my course introduced the concept, more or less. My professor also explained that one can look at the Cox model as a multiple logistic-regression model that considers both the occurrence of an event and the time until that event. Or something like that. Rosner says the same thing. So hopefully that's good enough?)