Kaplan-Meier analysis (KM analysis, also known as the Kaplan-Meier estimator or product-limit estimator) is a non-parametric statistical method for estimating the survival function from survival time data. It was introduced in 1958 by Edward L. Kaplan and Paul Meier and is today the standard method for visualising survival curves in clinical trials, particularly in oncology.
The product-limit estimator calculates the survival probability stepwise at each observed event (e.g. death, progression). Censored observations (participants for whom no event has occurred, e.g. at trial end or dropout) are included without biasing the result – provided the censoring is non-informative (non-informative censoring).
To compare two or more survival curves (e.g. treatment vs. control arm), the log-rank test (Mantel-Cox test) is used, which tests whether a statistically significant difference exists between the groups. In oncology trials, PFS and OS are routinely presented using Kaplan-Meier curves. The analysis is also established in cardiovascular trials (time to first MACE event) and other therapeutic areas. For CROs, correctly specifying the analysis in the Statistical Analysis Plan (SAP) is a key task.