Abstract
This study examines the effectiveness of administrative databases and national-level health surveys to increase our understanding of cancer health disparities among U.S. Pacific Islanders, a racially and ethnically diverse population that is medically underserved both in terms of screening and treatment. The study provides a systematic analysis of four data sources: 1) National Hospital Discharge Survey (NHDS), 2) Mortality Detail Files (MDF), 3) Surveillance, Epidemiology, and End Results (SEER) and 4) National Health Interview Survey (NHIS). These studies are examined to determine their efficacy in the statistical representation of Native Hawaiian and Pacific Islanders (NHPIs) in the United States. The paper presents a systematic and comparative evaluation of population coverage or sample attributes for each of the studies. The paper also evaluates the ability to obtain statistically robust measures of cancer incidence across select socioeconomic characteristics among the NHPIs in each study. The paper concludes with a summary of the strengths and limitations of these large-scale databases in advancing our understanding of cancer risks, prevalence, incidence, and mortality among NHPIs. Study findings will also examine the various ways that census data has been used in association with health data to provide detailed demographic, social, economic and geographic information on NHPIs. The strategic use of census data can allow us to in some ways overcome the lack of robust microdata on cancer patterns across NHPI subgroups and within geographic areas of specific concern for this population. The paper also discusses and evaluates some of the key barriers using administrative data and large-scale surveys to better understand cancer risks among NHPIs. Among these barriers, the lack of detailed information on NHPI subpopulations is particularly troublesome. While useful for establishing incidence, studies such as NHDS and MDFs continue to report only on Native Hawaiian, Samoan, and Guamanian/Chamorro subgroups and summarize the other NHPI populations in a residual other “Asian or Pacific Islander” category. In contrast, the SEER data base provides tremendous detail on NHPI subpopulations but does not attempt to estimate the annual population denominator estimates needed for the calculation of cancer rates. The paper concludes with a discussion of research methods and suggested approaches that may help overcome these kinds of data limitations and expand our understanding of cancer risks among this diverse and poorly understood population.
Citation Information: Cancer Epidemiol Biomarkers Prev 2010;19(10 Suppl):A2.