Which statement best describes interval censoring in survival analysis?

• A. The exact event time is known
• B. The event is not observed at all
• C. The exact event time is known to lie within an interval between observation times
• D. The event occurred before the study started

Answer: (C)

Interval censoring occurs when the exact time an event happens is not observed, but we can pin it down to lie somewhere between two scheduled observation times. In studies with periodic check-ins, you might learn the event has occurred at a visit, but the true moment it happened was after the previous visit and before the current one. The precise event time is unknown; it’s only known to lie within that interval. This is different from knowing the exact time, which would not be interval-censored; or from the event never being observed, which relates to right-censoring or missing data; or the event having occurred before the study began, which is left-censoring. Analyzing interval-censored data requires methods that account for the interval, such as Turnbull’s nonparametric estimator or interval-censored survival models, which use the probability that the event time lies within each observed interval.