---
title: "Kaplan-Meier Survival Curve"
slug: "kaplan-meier-survival-curve"
updated: 2026-10-23T07:07:18Z
published: 2026-10-23T07:07:18Z
canonical: "userguide.dataclassroom.com/kaplan-meier-survival-curve"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://userguide.dataclassroom.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Kaplan-Meier Survival Curve

The **Kaplan-Meier** curve displays time-to-event data. An '**event**' is some clearly defined happening that only happens once, such as death, component failure, or similar.

The curve shows the survival rate over time, which starts at 1 (100% at time zero) and falls as time goes on.

![](https://cdn.document360.io/62f8f41e-e041-42c4-b9bc-8ae1eb45d064/Images/Documentation/image-1782301200542.png)In this example, (see it [on DataClassroom here](https://dataclassroom.com/en/dataseteditor/49b68904-6e89-41f2-8866-a0be23416f32/graph/overlay/survival)) the Y axis shows 50% survival at about 73 weeks.

To display this curve, your dataset needs to be in this standardized format:

- A numerical column with the **Time to event** variable
- One row per event, where **Time to event** specifies when the event happened

Then you just need to select **Specialized graphs** -> **Survival curve** in the graphing tab:

![](https://cdn.document360.io/62f8f41e-e041-42c4-b9bc-8ae1eb45d064/Images/Documentation/image-1792430321140.png)

### Display options

You can display the curve optionally **with dots for the events**, with or without **gridlines**, and you can also choose to view the data **as a table**.

### Median and confidence interval

Beside the curve, you are also shown the calculated median survival time and 95% confidence intervals for that calculation. For the example above:

![](https://cdn.document360.io/62f8f41e-e041-42c4-b9bc-8ae1eb45d064/Images/Documentation/image-1782301580284.png)

The table view also shows these figures for each step-down in survival probability.

### Multiple groups

If you have an additional **categorical column** that **groups** the events/subjects into categories, you can display several curves at once:

![](https://cdn.document360.io/62f8f41e-e041-42c4-b9bc-8ae1eb45d064/Images/Documentation/image-1782301807970.png)

### Study drop-outs aka censoring

In such a survival study, there may well be participants or test subjects that leave the study before it ends. To take account of these, you need to add a **Status column** that marks whether the row signifies a standard study 'event' like a death, or whether it is a '**censoring**' event where the subject **left the study**.

The convention is that the status column be numeric with values **0 and 1**, with a **1 signifying an event**, and **0 meaning they were censored, i.e. left the study** at that time. So if no-one left the study, this column would be **all-ones**. The column should set to Categorical, but Numeric is also accepted.

*Note: The term 'censored' is also just a convention. It does not imply that anything is being deliberately hidden.*

You can optionally show these censoring events as small crosses on the curve:

![](https://cdn.document360.io/62f8f41e-e041-42c4-b9bc-8ae1eb45d064/Images/Documentation/image-1782302340261.png)

### Running statistical tests

You can run both a **Log Rank** and a **Cox Proportional Hazards** test for significant differences between the curves. Both of these tests are used to evaluate whether or not survival between the groups differs at any point along the curves. Use the **Run statistical tests** button underneath the table of medians:

![](https://cdn.document360.io/62f8f41e-e041-42c4-b9bc-8ae1eb45d064/Images/Documentation/image-1785772561725.png)

This opens a dialog box where you can see the results:

![](https://cdn.document360.io/62f8f41e-e041-42c4-b9bc-8ae1eb45d064/Images/Documentation/image-1785772677564.png)