Step-down curves and censored observations
A Kaplan-Meier survival curve shows the proportion of a cohort surviving (or event-free) over time as a step-down line. Each step down corresponds to an event (death, failure, or the outcome of interest). The curve does not decline at uniform intervals but only when an event occurs, reflecting the actual timing of data. Small vertical tick marks indicate censored observations: participants lost to follow-up or still alive at the end of the study whose true event time is unknown but the observation is still informative.
Comparing treatments and interpreting uncertainty
Two or more Kaplan-Meier curves on the same plot compare survival between treatment groups or cohorts. A log-rank test or Cox regression assesses whether the difference is statistically significant. The confidence interval around each curve (often shown as a shaded band) widens over time as the number of at-risk participants declines. The steep initial drop of a curve signals early events; a flat tail indicates stable, long-term survivors. Kaplan-Meier curves are standard in clinical trials, equipment reliability studies, and any analysis where time-to-event data is collected with follow-up dropout.