Which statement best describes how censoring is handled in survival analysis?

• A. Censored individuals contribute information up to their censoring time and are accounted for in the likelihood; non-informative censoring assumed
• B. Censored individuals contribute zero time
• C. Censored individuals are removed entirely from analysis
• D. Censoring always produces biased estimates

Answer: (B)

In survival analysis, censoring is handled by using all information available up to the time an individual is censored and by incorporating censored observations into the likelihood under the assumption that censoring is non-informative (independent of the outcome). This means censored individuals contribute person-time at risk up to their censoring time, and their exact event time is unknown; the model uses the probability of surviving beyond that censoring time. They are not removed from analysis and they do not contribute zero time. Proper handling with non-informative censoring means estimates can be unbiased, rather than automatically biased.