Interpreting Log-Rank Test Results
You'll learn: How to read p-values and chi-squared statistics from a log-rank test to decide whether survival curves differ meaningfully.
After computing a log-rank test comparing two or more survival curves, you need to interpret the output to draw conclusions. The test produces two key numbers: a chi-squared statistic and a p-value.
Understanding the Chi-Squared Statistic
The chi-squared statistic measures how much the observed event counts in each group deviate from what we'd expect if all groups had identical survival patterns. A larger chi-squared value indicates greater difference between the survival curves. Think of it like measuring the distance between two paths—the farther apart they are, the bigger the number.
Reading the P-Value
The p-value tells you the probability of seeing differences this large (or larger) purely by chance if the groups truly had the same survival. It's your decision-making tool:
- p < 0.05 (common threshold): Strong evidence that survival curves differ significantly. Reject the null hypothesis—the groups have different survival experiences.
- p ≥ 0.05: Insufficient evidence to claim the curves differ. Cannot reject the null hypothesis—observed differences might just be random variation.
Practical Example Interpretation
Suppose you compare survival between two treatment groups and get chi-squared = 8.3, p = 0.004. You'd conclude: "The survival curves differ significantly (p = 0.004), suggesting the treatments have different effects on survival time."
If chi-squared = 1.2, p = 0.27, you'd say: "No significant difference in survival was detected between groups (p = 0.27)."
Key Takeaway: A log-rank test p-value below your significance threshold (typically 0.05) means the survival curves differ significantly; the chi-squared statistic quantifies the magnitude of that difference.