Browse topics

Survival analysis, Kaplan-Meier curves and hazard ratios

Censoring, the Kaplan-Meier product-limit method, the log-rank test and Cox hazard ratios, with a worked conversion from a hazard ratio to an absolute benefit and NNT.

Tier 1Medical sciences concept

Written with AI assistance and fact-checked against the sources below. For exam study, not clinical decisions. How notes are made

3 questions 8 flashcards

At a glance

Time-to-event data record whether and when an event happens. Some participants are censored: their follow-up ends before an event.
Kaplan-Meier (product-limit) estimates the survival curve, using everyone for as long as they are observed.
Log-rank test compares curves and gives a p-value only.

Loading the rest of the note…

Finished reading?

Test yourself on it:3 questions 8 flashcards

Sources

Statistics notes: Time to event (survival) dataBMJ (Altman and Bland, 1998;317:468), 1998
Statistics notes: Survival probabilities (the Kaplan-Meier method)BMJ (Bland and Altman, 1998;317:1572), 1998
Statistics notes: The logrank testBMJ (Bland and Altman, 2004;328:1073), 2004
Hazards and hazard ratios in clinical trialsAntimicrobial Agents and Chemotherapy (Spruance et al., 2004;48:2787-92), 2004
Dapagliflozin in patients with heart failure and reduced ejection fraction (DAPA-HF)N Engl J Med (McMurray et al., 2019;381:1995-2008), 2019

Last updated: 14 November 2026

What changed

14 November 2026 Turnstile bot check on sign-in and signed-out feedback