Glossary

Kaplan-Meier Curve

A Kaplan–Meier curve is a non-parametric plot that estimates the survival function from lifetime data. It displays the probability of surviving beyond a given time as a step function, with downward steps occurring at each observed event time. The curve naturally handles censor...

Definition

A Kaplan–Meier curve is a non-parametric plot that estimates the survival function from lifetime data. It displays the probability of surviving beyond a given time as a step function, with downward steps occurring at each observed event time. The curve naturally handles censored observations — individuals who have not yet experienced the event by the end of the study.

Why It Matters

In survival analysis, simply calculating the proportion of events would ignore the fact that some participants are censored and have not yet had the opportunity to experience the outcome. The Kaplan–Meier method uses all available information, including censored cases, to produce an unbiased estimate of survival probability over time. It is the standard visualisation in clinical trials, reliability engineering, and any study of time-to-event data.

Example

An oncology trial compares two chemotherapy regimens. The Kaplan–Meier curve for the new drug lies consistently above the curve for standard care. At 12 months, the survival probability is 78% for the new drug versus 65% for standard care. The curves are accompanied by tick marks indicating censored patients, and a log-rank test p-value of 0.03 confirms the difference is statistically significant.

Related Terms

Software Notes

  • SPSS: Analyze > Survival > Kaplan-Meier. Define the time variable and status indicator (event = 1, censored = 0). Use the Compare Factor Levels option to run log-rank tests between groups. Request survival tables and plots under Options.
  • R: survival::survfit(Surv(time, status) ~ group, data = df) fits the Kaplan–Meier model. survminer::ggsurvplot(fit, pval = TRUE, risk.table = TRUE) produces publication-ready curves with p-values, confidence intervals, and risk tables.
  • Stata: stset time, failure(status) declares survival data. sts graph, by(group) plots Kaplan–Meier curves by group. sts test group performs the log-rank test. sts list displays the survival table.