ABSTRACT
Epidemiological and biomedical studies often require modeling and analysis for time-toevent data, where a subject is followed up to an event (e.g., death or onset of a disease) or is “censored,” whichever comes first. Survival models (see, e.g., Cox and Oakes [1]) are widely used in biostatistics and epidemiology for analyzing time-to-event data, including, perhaps, several censored observations. If the event does not occur for a subject during the period of the study, then the subject’s time to event is censored at the study end point. This situation is described as “right censoring” and is, perhaps, the most common. Analogously, certain study designs can produce “left-censored” or “interval-censored” data. As opposed to modeling disease incidence and mortality, survival modeling focuses on how many are expected to survive after a certain period of time, the rate of failure, and the factors driving shortened or prolonged survival; all of these may be influenced by several factors such as gender, race, age, type of cancer, treatment obtained, and access to health care facilities.
