Melanoma prognosis in the United States: Identifying barriers for improved care.

Abstract

Despite improvements in melanoma mortality, disparities in melanoma survival persist. We evaluated possible sociodemographic and health care-based predictors of differences in melanoma survival in the United States by using the melanoma mortality-to-incidence ratio (MIR). State-based MIRs were calculated by using US cancer statistics data from 1999 to 2014. Pearson correlations and linear regressions were used to determine associations between MIR and dermatologist density, primary care provider density, number of physicians by state, number of National Cancer Institute-designated cancer centers, health care spending per capita, average household income, racial/ethnic makeup of the population, percentage of uninsured individuals, and percentage with a bachelor's degree. The mean overall MIR was 0.15 ± 0.04; only Alaska was an outlier (0.24). No state MIRs increased significantly over time; MIR decreased for most states. Multivariable analysis revealed that states with more active physicians (P = .02) and a higher percentage non-Hispanic whites (P = .004) had higher MIRs (poorer survival). Significant Pearson correlations were seen between MIR and melanoma incidence (r = -0.72, P < .001), melanoma mortality (r = 0.38, P < .001), dermatologist density (r = 0.32, P < .001), and National Cancer Institute-designated cancer center count (r = -0.12, P = .001). Melanoma survival is improved in higher-incidence areas and areas with higher dermatologist density. These findings highlight areas of poorer melanoma survival and the need for local studies evaluating disparities in melanoma survival.