Abstract

APPLICATIONS OF THE KAPLAN–MEIER ESTIMATOR IN THE MEDICAL FIELD: A NARRATIVE REVIEW

Dr. S. Kumar*, Dr. C. Sankar

ABSTRACT

The Kaplan–Meier (KM) estimator, introduced in 1958 as a non-parametric method for estimating the survival function from time-to-event data, remains one of the most widely applied statistical tools in clinical and biomedical research. Its ability to handle right-censored observations without assuming an underlying distribution has made it the default first step in survival analysis across virtually every branch of medicine. This narrative review synthesises recent literature (predominantly 2023-2026) describing applications of the KM estimator in oncology, cardiology, nephrology, infectious disease (tuberculosis and HIV), critical care and sepsis research, endocrinology and diabetes, pharmacovigilance and pharmacoepidemiology, transplantation medicine and emerging machine-learning-based survival modelling. We alsosummarise the mathematical basis of the estimator, its common companion methods (the log-rank test and Cox proportional hazards regression) well-documented limitations such as bias under competing risks and instability with sparse late follow-up and current directions including differentially private and deep-learning-based extensions. The review is intended as a practical reference for clinicians, pharmacy practitioners and biostatistics trainees who encounter Kaplan–Meier curves in the medical literature.

Keywords: Kaplan–Meier estimator; survival analysis; time-to-event data; censoring; log-rank test; Cox proportional hazards model; medical statistics.