Kaplan Meier Survival Analysis

The Kaplan-Meier calculator estimates survival probability over time from time-to-event data, with proper handling of censored observations. Supports four analysis modes from simple single-cohort curves to propensity-score-matched comparisons.


Step 1 — Choose analysis type

Type Use case
Single Group Analysis Descriptive survival in one cohort
Two Group Comparison Treatment vs control with log-rank test
Multiple Group Comparison Multi-arm trials with pairwise comparisons
Propensity Score Matching Observational studies with confounding

Step 2 — Provide your data

Each subject needs:

  • Time — follow-up time
  • Event — 1 (event occurred) or 0 (censored)
  • Group — group label (for multi-group modes)
  • Covariates — for propensity matching

Supports Excel Import, Manual Entry, and Sample Data Generator.


Step 3 — Results

Always shown:

  • Survival points table — time, n at risk, events, censored, S(t), SE, 95% CI
  • Median Survival Time with 95% CI
  • Mean Survival Time with 95% CI
  • Kaplan-Meier Curve with censored marks

For group comparisons:

  • Log-rank test — chi-square, df, p-value
  • Hazard Ratio with 95% CI
  • Pairwise comparisons (multi-group)

For propensity matching:

  • Propensity scores and matching results
  • Before/after balance assessment
  • Matched-pair analysis

A UniversalChatBot is available for discussion.


Statistical methods used

Kaplan-Meier product-limit estimator

S(t_i) = S(t_{i-1}) * (1 - d_i / n_i)

Standard error — Greenwood's formula

SE(S(t)) = S(t) * sqrt(sum(d_i / (n_i * (n_i - d_i))))

Median survival — smallest t where S(t) <= 0.5, with linear interpolation.

Log-rank test

For each event time: observed vs expected events per group.

chi2 = (O - E)^2 / V with df = k-1.

Hazard ratio from log-rank

HR = exp((O - E) / V) with 95% CI.

Propensity Score Matching — nearest-neighbour 1:1 matching on logistic regression propensity scores.