The Selection tab is used to select the records to be included in the analysis. On the Selection tab for a Survival session, there are checkboxes for standard case selections made for survival runs. In addition, selection statements may be created in the Case Selection box to reduce the number of records included in an analysis based on specific variables in the case data files. If no selection statements are made, all records in the database will be included.
When making selections, start at the top and work your way down the window.
The Standard Case Selections box contains a set of case selection or exclusion criteria commonly associated with survival analyses. Each standard selection or exclusion can be accepted or rejected by marking or clearing its checkbox. When a new Survival Session is started, some of the standard selections or exclusions will be automatically marked. These default selections represent the standard selections most commonly used for a survival analysis.
Select this option to include all cases that were microscopically confirmed. The case will be included in the analysis if the Diagnostic Confirmation variable indicates that there was microscopic confirmation of the morphology of the cancer.
Select this option to include only cases that were actively pursued for follow-up information. If the Type of Follow Up variable indicates that the case was actively followed, the case will be included in the analysis.
Select this option to include only cases with a malignant behavior. The Behavior variable is part of the morphology of cancer. The SEER program only collects information on in situ and malignant cancers. If the behavior code is malignant, the case will be included in the analysis.
Select this option to exclude all cases where cancer was reported through only a death certificate or autopsy. If the Type Of Reporting Source variable indicates that a death certificate or autopsy was the only source of information for the case, the case will be excluded from the analysis.
Select Exclude All Second and Later Primaries to exclude all case records that are second or later primaries. The Sequence Number variable indicates the chronology of diagnosis of all reportable malignant and/or in situ cancers over the known history of the patient. The case will be excluded from analysis if it is not the first primary cancer.
Select Exclude All Records for Persons with Multiple Primaries to exclude all case data records that are one of multiple primaries. Only case records that are the first and only primary for a patient are kept. The Sequence Number variable indicates the chronology of diagnosis of all reportable malignant and/or in situ cancers over the known history of the patient. The case will be excluded if it is NOT the first and only primary cancer.
Exclusions required to match Expected survival data are always made when relative survival or crude probability of death using expected survival are calculated. Records with values not found in the expected survival data must be excluded. For example, there are no expected survival data for unknown race. Therefore, case records with unknown race are excluded when using expected survival data in calculations. When you calculate survival statistics that do not use expected survival data (such as observed survival), you have the option to turn these exclusions on or off. Using these exclusions for observed survival allows you to analyze the same cohort that would be used in relative survival calculations -- assuming you are working with the same expected survival file.
Select this option to exclude all cases with ages outside the boundaries of the expected survival table or with unknown ages.
Select this option to exclude all cases with an unknown age date (the date at which the age was coded). In SEER, the age date is the date of diagnosis.
Select this option to exclude all cases with a missing or unknown value for any remaining expected survival table variable (excluding age and year when age was coded). For SEER*Stat-supplied expected survival tables, the only "other" variable is race.