In survival analysis, when a participant withdraws from a study, how is this usually handled?

• A. They are counted as deaths.
• B. They are censored at the time of withdrawal.
• C. They remain in the risk set for the entire study.
• D. They are dropped from analysis and not counted in any way.

Answer: (B)

In survival analysis, when a participant withdraws, their data are censored at the time of withdrawal. This means we know they did not experience the event up to that moment, but what happens after withdrawal is unknown. The observation contributes to the analysis up until withdrawal, and they drop out of the risk set at that exact time. They are not counted as a death, and they are not completely excluded from the analysis—only their subsequent fate is unknown. This approach preserves information up to withdrawal and allows proper estimation of survival probabilities under non-informative censoring.