Focus on Survivorship: Refining Complete Prevalence Estimates Using Local Cancer Registry Data

Purpose: This study applies NCI's ComPrev software to ten years of North Carolina Central Cancer Registry (NCCCR) data to demonstrate the added value of using complete prevalence (CP) methods to estimate cancer prevalence compared with the prior standard, limited-duration prevalence (LDP). Methods: All cases for years 1995-2004 were obtained from the NCCCR and merged with vital status and date of death from the NC State Center for Health Statistics' vital statistics to form the “case data.” Population data for four expanded races and individual ages, originally from the U.S. Census Bureau, were obtained from the SEER program. Case and population data were integrated using SEER*Prep software. SEER*Stat software was used to calculate 10-year LDP based on first primary cancer as of January 1, 2005. ComPrev software was subsequently used to convert LDP into CP for each cancer site, gender, and race combination. CP applies a completeness index based on incidence and survival to adjust for underascertainment of cases due to limited years of surveillance, providing a more accurate estimate of survivor burden. Results: LDP methods estimated 228,457 cancer survivors living in NC as of January 1, 2005. CP methods estimated 371,537 individuals living with cancer, 63% more than LDP estimates. Compared to cancers with typically short survival duration or later age at diagnosis, cancers of greater survival duration or earlier age at diagnosis tended to have greater difference between LDP and CP estimates. Breast, prostate, colorectal, and melanoma comprise 205,747 (55%) of the cancer survivors in NC as estimated by CP methods. Conclusions: The number of cancer survivors in the US is expected to grow from the 2006 estimate of >11 million, and more refined prevalence estimates are needed for survivorship planning. In North Carolina, the commonly used method of LDP based on SEER data underestimates the survivor burden by nearly 40%. Regional, gender-based, and racial differences suggest local programs should understand their community's prevalent cancer population when prioritizing local survivorship services. Future analysis will use MIAMOD methods as a supporting method of estimating CP, and updated case data to generate January 1, 2008 prevalence estimates. This abstract is one of the 17 highest scoring abstracts of those submitted for presentation at the 34th Annual Meeting of the American Society of Preventive Oncology, to be held March 20-23, 2010 in Bethesda, MD.