Log Rank Test

The log-rank test is a nonparametric hypothesis test that compares survival curves between two or more groups by contrasting observed and expected event counts at each event time.

로그순위 검정은 각 사건 시점에서 관측 사건 수와 기대 사건 수를 대조하여 두 개 이상 집단의 생존곡선을 비교하는 비모수 가설검정이다.

How do I test whether survival curves differ between groups?

집단 간 생존곡선이 다른지 어떻게 검정하나요?

The log-rank test compares time-to-event distributions between two or more groups and accounts for censored observations. It is usually reported with Kaplan-Meier curves and has the most power when the hazards in the groups are proportional.

When to use Log Rank Test

• Time-to-event data with censoring and a grouping variable
• Comparing Kaplan-Meier curves for treatment arms or segments
• An unadjusted comparison without covariates

When to use something else

• Adjusting for covariates; use the Cox proportional hazards model
• Survival curves that cross or early differences matter; use the Wilcoxon (Breslow) test or restricted mean survival time

Assumptions and requirements

• Non-informative censoring
• Independent observations between groups
• Most powerful under proportional hazards

What the analysis reports

• Chi-square statistic with degrees of freedom
• p-value
• Observed and expected events per group
• Kaplan-Meier curves
• Median survival time by group

How to report it

Survival differed significantly between groups, log-rank χ²([df]) = [value], p = [value].

Related methods

• Kaplan-Meier Estimator
• Cox Proportional Hazards Model
• Restricted Mean Survival Time

Run Log Rank Test on Quorant