Checklist for NPCR and SEER Incidence – U.S. Cancer Statistics Public Use Database, 2001–2016 Analyses
The database includes variables that can be used to restrict analyses to the states meeting U.S. Cancer Statistics publication criteria during the most commonly analyzed multi-year time periods, specifically—
- All years of data in the database (variable USCS0116 for diagnosis years 2001–2016).
- The most recent 10 years of data (USCS0716 for diagnosis years 2007–2016).
- The most recent 5 years of data (USCS1216 for diagnosis years 2012–2016).
If you are conducting a multi-year analysis and want to restrict it to the states that met publication criteria during each of the years, did you use variable USCS0116, USCS0716, or USCS1216 and also use the Year of Diagnosis variable to restrict to the corresponding year range on the SEER*Stat Selection tab?
- This is important for trend analyses, as the same states need to be included for each year being analyzed for comparisons.
- The Year of Diagnosis variable is used in combination with the predefined USCS variable to exclude the non-relevant years. For example, if USCS1216 is used, then Year of Diagnosis should also be restricted to diagnosis years 2012–2016 in the SEER*Stat Selection tab.
- If you would like to analyze a range of years other than those predefined variables, please contact CDC at email@example.com and we will create a new variable for you.1
If you are analyzing just one year of data, did you use the variable USCS Standard and restricted the analysis to the specific Year of Diagnosis in the SEER*Stat Selection tab?2
Common Selection and Reporting Considerations
- If you are reporting state-level race, ethnicity or race/ethnicity combinations, have you suppressed data from the registries that opted out of reporting these data items? Race and ethnicity combinations can be excluded using the state race eth suppress variable; race-only or ethnicity-only suppressions should be done manually in the SEER*Stat Selection tab.3
- If a user-defined primary site variable was created (rather than using the Site recode ICD-O-3/WHO 2008 variable)—
- Did you exclude leukemias and lymphomas (9590–9992)?
- Did you consider excluding Kaposi sarcoma (9140) and mesothelioma (9050–9055)?4
- If your analysis includes histology, and if appropriate for the cancer site, did you use the Diagnostic Confirmation variable to specify the analysis be limited to Microscopically confirmed cases?5
- If you are analyzing sex-specific cancers (such as prostate cancer or female breast cancer), did you limit the analysis to the appropriate sex to get the correct population denominator?6
- When reporting rates, have you included the label “per 100,000 persons,” “per 100,000 women,” or “per 100,000 men”?
- Have you included citations for the—
- Percentage of United States population coverage provided by the database?
- NPCR and SEER Incidence – U.S. Cancer Statistics 2001–2016 Public Use Research Database?7
1See data dictionary pdf icon[PDF-855KB] – USCS0116, USCS0716, and USCS1216 variable descriptions.
2See data dictionary pdf icon[PDF-855KB] – USCS Standard variable description.
3See data dictionary pdf icon[PDF-855KB] – state race eth suppress, Race Recode, Origin recode NHIA, and NHIA derived Hisp origin variable descriptions.
4See data dictionary pdf icon[PDF-855KB] – Cautionary Notes section entitled Primary Site Variables.
5See data dictionary pdf icon[PDF-855KB] – Diagnostic Confirmation variable descriptions.
6See data dictionary pdf icon[PDF-855KB] – Sex variable description.
7See data dictionary pdf icon[PDF-855KB] – Data Citation section.