Using Aggregated Data (SURVIVAL command)

When aggregated survival information is available, the number of censored and uncensored cases at each time point must be entered. Up to two records can be entered for each interval, one for censored cases and one for uncensored cases. The number of cases included on each record is used as the weight factor. If control variables are used, there will be up to two records (one for censored and one for uncensored cases) for each value of the control variable in each interval. These records must contain the value of the control variable and the number of cases that belong in the particular category as well as values for survival time and status.

Example

DATA LIST  / SURVEVAR 1-2 STATVAR 4 SEX 6 COUNT 8.
VALUE LABELS  STATVAR 1 'DECEASED' 2 'ALIVE'
              /SEX 1 'FEMALE' 2 'MALE'.
BEGIN DATA
 1 1 1 6
 1 1 1 1
 1 2 2 2
 1 1 2 1
 2 2 1 1
 2 1 1 2
 2 2 2 1
 2 1 2 3
   ...
END DATA.
WEIGHT COUNT.
SURVIVAL TABLE = SURVEVAR BY SEX (1,2)
  /INTERVAL = THRU 10 BY 1
  /STATUS = STATVAR (1) FOR SURVEVAR.
  • This example reads aggregated data and performs a SURVIVAL analysis when a control variable with two values is used.
  • The first data record has a code of 1 on the status variable STATVAR, indicating that it is an uncensored case, and a code of 1 on SEX, the control variable. The number of cases for this interval is 6, the value of the variable COUNT. Intervals with weights of 0 do not have to be included.
  • COUNT is not used in SURVIVAL but is the weight variable. In this example, each interval requires four records to provide all of the data for each SURVEVAR interval.