Comparing Multiple Survival Curves
You'll learn how to visually and descriptively compare survival experiences across different groups in your data.
What This Means
Once you've calculated Kaplan-Meier survival curves for different groups (called strata)—like treatment vs. control, male vs. female, or different product tiers—you need to compare them to understand whether survival differs meaningfully between groups.
This involves two main activities: visualization and description.
Visualizing Multiple Curves
The most powerful approach is plotting multiple survival curves on the same axes. Each stratum gets its own curve, typically color-coded or styled differently. You can immediately see:
- Which group has better survival (curve is higher = more survivors at each time point)
- When differences emerge (curves may start together then diverge)
- Size of the gap between groups over time
Think of it like comparing runners in a race—plotting their completion rates over time shows who performs better and when they pull ahead.
Describing the Differences
Beyond the plot, you describe differences numerically:
- Median survival times for each group (from lesson 816)
- Survival probabilities at key time points (e.g., 1-year survival rate: 75% vs. 60%)
- Confidence intervals to show uncertainty in each curve
- Visual separation of confidence bands (non-overlapping suggests real differences)
For example: "The treated group shows 80% survival at 12 months compared to 65% in the control group, with non-overlapping confidence intervals suggesting a meaningful difference."
Why It Matters
Comparing curves answers critical questions: Does the new treatment work better? Do customer segments behave differently? Are certain risk factors associated with worse outcomes?
Key Takeaway: Comparing multiple survival curves uses overlaid plots and descriptive statistics to reveal whether and when different groups experience different survival patterns.