Kaplan-Meier Survival Estimator
Also known as: product-limit estimator, km curve, kaplan-meier sağkalım analizi
The Kaplan-Meier estimator, introduced by Kaplan and Meier in 1958, is a non-parametric method that estimates the survival curve — the probability of remaining event-free over time — from right-censored time-to-event data. The log-rank test is the companion procedure used to compare survival curves between groups.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
+12 more
When to use it
Kaplan-Meier is the right tool when you have time-to-event data with right-censoring and the censoring is non-informative — that is, subjects censored at a given time carry the same future risk as those still under observation. Use it to describe a single survival curve, or to compare curves across groups, in which case the log-rank test assesses whether the separation between curves is statistically meaningful. It does not adjust for covariates; when you need to model the effect of explanatory variables, Cox regression is the appropriate extension.
Strengths & limitations
- Handles censored observations directly, retaining their information instead of discarding it.
- Produces a survival curve that is visual and easy to interpret.
- A universal standard in health research, immediately recognised by reviewers.
- Cannot adjust for covariates; multivariable analysis requires Cox regression.
- Confidence intervals become wide at the tails of the curve in small samples.
- Assumes non-informative censoring; if censoring is related to prognosis, estimates are biased.
Frequently asked
What is censoring?
Censoring occurs when you do not observe the event for a subject within the study — for example because the study ends or the subject drops out. With right-censoring you know the subject was event-free up to a certain time but not beyond. Kaplan-Meier keeps these subjects in the risk set up to their censoring time rather than discarding them, which avoids the bias of treating censored times as events.
How does Kaplan-Meier differ from Cox regression?
Kaplan-Meier is a non-parametric estimator that describes and compares survival curves but cannot adjust for covariates. Cox regression models how explanatory variables affect the hazard, so it is used when you need multivariable adjustment or to quantify the effect of predictors. The two are complementary: Kaplan-Meier for description, Cox for modelling.
What is median survival time?
The median survival time is the time at which the estimated survival probability first falls to 0.5 — the point by which half of the subjects are expected to have experienced the event. It is the standard summary of a Kaplan-Meier curve and should be reported with its 95% confidence interval.
How do I compare two survival curves?
Use the log-rank test, which compares the observed and expected number of events across groups over time. A separation between curves is meaningful only when the log-rank test confirms it and the proportional-hazards assumption holds.
Sources
- Kaplan, E. L. & Meier, P. (1958). Nonparametric Estimation from Incomplete Observations. Journal of the American Statistical Association, 53(282), 457–481. DOI: 10.1080/01621459.1958.10501452 ↗
- Kleinbaum, D. G. & Klein, M. (2012). Survival Analysis: A Self-Learning Text (3rd ed.). Springer. ISBN: 978-1441966452
How to cite this page
ScholarGate. (2026, June 1). Kaplan-Meier Survival Estimator. ScholarGate. https://scholargate.app/en/survival/kaplan-meier
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Log-Rank TestSurvival↔ compare
- Nelson-Aalen EstimatorSurvival↔ compare