Survival Function table for Fit Cox Model with Fixed Predictors only

The Survival Function table provides the probability that a unit survives until a particular time. The survival function in the table is for an individual with the continuous predictors at their mean values and categorical predictors at their reference levels. Use the table to get estimates of the survival function for the settings of the predictors in the table. If the model includes strata, then the results include a separate survival function for each stratum. Compare the survival functions for different strata to estimate the effect of the strata.

If the model fits the data poorly, the estimates of the survival function can be inaccurate.

When you run the analysis, you can specify the reference level for categorical predictors. You can also specify predictor values to show on a survival plot so that you can compare the survival functions with different settings of the predictors.

Time

The time until a predicted probability of units survives. The values in the table are times in the data that cover the study period.

Number at Risk

The number of units at risk equals the number of units that have entered the study but have not experienced the event of interest. The number of units at risk for a given time differs from the number of units in the entire study for two reasons.

Units can enter the study at any time. If a unit did not enter the study yet, then the unit is not at risk.
Units that already experienced the event are not at risk anymore.

Event count

The number of units that experience the event at the time for the row. Censored observations do not add to the event count.

Survival probability

The chance that a unit with the combination of predictor settings for the function survives until the time for the row.

Standard error

The standard error estimates the variation in the survival function for the predictor settings for the table and the time for the row. Use the standard error to measure the precision of the estimate of the survival function. The smaller the standard error, the more precise the estimate of the survival function. A more precise estimate produces a narrower confidence interval.

CI

The confidence interval for the survival function provides a range of likely values for the survival function for the predictor settings for the function and the time for the row.