Calculation of complete cancer prevalence requires several years of incidence data and accurate vital status information at end of follow-up.
Five states from the Surveillance, Epidemiology, and End Results (SEER) program (http://seer.cancer.gov) have cases diagnosed from 1975 and allow estimation 30-year limited duration prevalence which is close to complete.
Fewer years of cancer incidence are available from states funded by SEER and/or the CDC National Program for Cancer Registries (NPCR) (http://www.cdc.gov/Cancer/npcr/).
Gaps in data collection prevent the calculation of cancer prevalence for many states. A statistical model is used that predicts cancer prevalence1 from state specific cancer mortality data from National Center of Health Statistics (NCHS) and cancer survival model2 adjusted to represent cancer survival in each respective state. The method has been validated against reported incidence cases for 37 states from SEER and NPCR.
1. De Angelis R, Tavilla A, Verdecchia A, Feuer EJ, Mariotto AB. Breast cancer survivors in the US States: geographical variability and time trends 2005-2015 Submitted to Cancer.
2. Mariotto A, Capocaccia R, Verdecchia A, Micheli A, Feuer EJ, Pickle L et al. Projecting SEER Cancer Survival Rates to the US: An Ecological Regression Approach. Cancer Causes and Control 2002; 13(2):101-111.