Geographic variations of racial/ethnic disparities in cervical cancer mortality in Texas.

OBJECTIVES: To examine how racial/ethnic disparities of cervical cancer mortality vary geographically and to identify factors contributing to the variation.
METHODS: Using the population-weighted risk difference, the authors investigated geographic patterns of racial/ethnic disparities in cervical cancer mortality in Texas based on data from 1995 to 2008 georeferenced at the census tract level. In addition, we considered the impact of seven factors--stage at diagnosis, spatial access to health care, and five factors that were created from available demographic data: socioeconomic status (SES), the sociodemographic factor, the percentage of African Americans, the health insurance factor, and the behavioral factor--on racial/ethnic disparities in the analysis using multivariate logistic regression.
RESULTS: SES, the sociodemographic factor, the percentage of African Americans, and racial/ethnic disparities in late-stage diagnosis in a census tract were independent predictors of a census tract's displaying significant racial/ethnic disparities in cervical cancer mortality. Compared with a census tract with the highest SES, a census tract with the lowest SES was more likely to have higher mortality rates in African Americans (odds ratio 4.19, confidence interval 2.18-8.07) or Hispanics (odds ratio 8.15, confidence interval 5.27-12.61) than non-Hispanic whites after adjusting for covariates. Health insurance expenditures also influenced racial/ethnic disparities in mortality, although this effect was attenuated after adjusting for covariates. Neither our calculated behavioral factor nor spatial analysis of access to health care explained racial/ethnic gaps in mortality.
CONCLUSIONS: Findings from this study could allow cervical cancer intervention programs to more clearly identify areas that would reduce disparities in cervical cancer outcomes.