Predictors of neighborhood risk for late-stage melanoma: addressing disparities through spatial analysis and area-based measures.

Minority populations have disproportionately more advanced stage melanoma and worse survival. To clarify the impact of race and ethnicity on late-stage melanoma diagnosis, we performed spatial analysis of geocoded melanoma cases diagnosed in Florida, 1999-2008, to identify geographic clusters of higher-than-expected incidence of late-stage melanoma and developed predictive models for melanoma cases in high-risk neighborhoods accounting for area-based poverty, race/ethnicity, patient insurance status, age, and gender. In the adjusted model, Hispanic ethnicity and census tract-level poverty are the strongest predictors for clustering of late-stage melanoma. Hispanic whites were 43% more likely to live in neighborhoods with excessive late-stage melanoma (P<0.001) compared with non-Hispanic whites (NHW). For every 1% increase in population living in poverty, there is a 2% increase in late-stage melanoma clustering (P<0.001). Census tract-level poverty predicted late-stage melanoma similarly among NHW and Hispanic whites. The impact of insurance coverage varied among populations; the most consistent trend was that Medicaid coverage is associated with higher odds for late-stage melanoma. The finding that Hispanics are most likely to reside in high-risk neighborhoods, independent of poverty and insurance status, underscores the importance of addressing, and overcoming community-level barriers to melanoma care.