Recorded presentation

Paul Stehr-Green, DrPH, MPH¹, Emily Puukka, MS¹, Thomas M Becker, MD, PhD², and Lori Lambert, MA². (1) Northwest Portland Area Indian Health Board, Northwest Tribal Epidemiology Center, 527 SW Hall Street #300, Portland, OR 97201, 503.228.4185, epuukka@npaihb.org, (2) Department of Public Health & Preventive Medicine, Oregon Health Sciences University, 3181 SW Sam Jackson Park Road, CB669, Portland, OR 97201-3098

Cancer incidence among American Indians and Alaska Natives (AI/ANs) is underestimated due to racial misclassification in public health data sets. We estimated current cancer incidence for AI/ANs living in the Pacific Northwest using probabilistic matching algorithms to link Indian Health Service enrollment data to the Idaho, Oregon, and Washington state cancer registries. For 1996 to 1999, these linkages identified 423 cancer cases that were not previously recognized in the state registries to have occurred among AI/ANs (an increased ascertainment of 89.6%), resulting in a total of 895 cases; by the time of this presentation, we will include two additional years of data (2000-2001) in these analyses. The overall age-adjusted cancer incidence rate among AI/ANs (269 per 100,000 person-years [pyr]) was approximately double that estimated prior to the linkage. For common cancer sites, post-linkage rates among AI/ANs (87 and 78 per 100,000 pyr for breast and prostate cancers, respectively) were also higher than had been estimated prior to the linkages. In Idaho, where statewide survival data are available, preliminary results indicate an unadjusted 5-year survival of 45.4%, with an unadjusted mean survival of 6.8 years among AI/ANs for all cancer sites combined. The post-linkage cancer incidence estimates reflect more complete data, and indicate that cancer among AI/ANs is a public health problem of greater magnitude than was previously appreciated. Our findings demonstrate the utility of using probabilistic data linkages to improve cancer rate estimates that can be used to inform priority-setting and program management for cancer education, prevention, and control.

Learning Objectives: At the conclusion of the session, the participant will be able to

• Articulate the major steps involved with conducting a record linkage.
• Discuss the impact of racial misclassification on cancer incidence estimates.
• Recognize disparities in the burden of cancer among AI/ANs versus non-AI/ANs.

Keywords: American Indians, Cancer

Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.

Recorded presentation