Evidence-appraisal glossary
Censoring
Censoring happens in survival studies when a participant's exact time to the event of interest is unknown because the study ended, they dropped out, or they were lost to follow-up before it occurred. We only know they stayed event-free up to their last contact.
Also called: Right-censoring, Censored data, Loss to follow-up.
Survival analysis measures time until an event such as death, relapse, or discharge. For some participants that event never occurs during the study window, or they leave before it does. Rather than discard them, analysts censor them: they contribute follow-up time up to their last known event-free moment, then exit the risk set. This is why Kaplan-Meier curves and hazard ratios can use partial information.
When reading a study, check how much censoring occurred and why. Right-censoring (the usual kind) is safe only when it is non-informative, meaning dropping out is unrelated to prognosis. If sicker patients withdraw or are lost to follow-up, censoring becomes informative and can bias the estimated survival, usually making treatment look better than it is. Look for the number lost to follow-up, whether it differs between arms, and any sensitivity analysis testing the censoring assumption. Heavy or unbalanced censoring should lower your confidence in the reported curves.
This is a plain-language methodology definition for reading research. It is general education, not medical advice.