KMT01 [Spec Added]

### Table Name

KMT01

### Description

Kaplan-Meier (KM) Event-free Rates over Time

### Analysis dataset

ADTTE

### Analysis filtering

Will be controlled by the user

### Analysis population

`IT`

### Column Variable

`ARM`

### Analysis variables

`AVAL`, `CNSR`

### Analysis concept


### Statistics

Allow α level for CIs to be adjusted, default is 2-sided 5%.
Use proc lifetest to obtain time to event information. Event free-rates are presented as percentages, i.e. the KM estimates from PROC LIFETEST need to be multiplied by 100.
As an option, display p-values for differences in event-free rates between treatment arm(s) and control group at the various timepoints. P-values are from pairwise comparisons for event free rates at x-days/months/years of the various treatment groups with control. These are calculated using a z−test with standard errors for the KM-estimates computed via the Greenwood method.
As an option, display KM estimates of pairwise differences in event-free rates and corresponding 95% CIs. For pairwise comparisons using the difference in event-free rates, the value for the control group is the subtrahend.
Patients at risk at time t include those patients who have not had an event before time t, and who were not censored. Patients who have an event at or after time t are deemed at risk
For consistency with SAS PROC LIFETEST output, time points after the maximum observed time will not be displayed since there is no information provided from PROC LIFETEST for later time points. There is an option to display the last observation (for each group) (p-values are not displayed for the last observation).

### Default treatment of zero count rows

Remove zero count rows by default

### Additional comments

Allow the following sections of the table to be optional: Number of events, number of censored cases, KM estimates of pairwise differences in event-free rates and corresponding 95% CI, p-values.
Column Timepoint: User must be able to specify the timepoints to be displayed, including the units (e.g. weeks or months instead of years)