Disparities in Overall Survival in Patients with Melanoma by Race/Ethnicity, Socioeconomic Status, and Healthcare Systems

1 Department of Dermatology, Kaiser Permanente Los Angeles Medical Center, Los Angeles, CA 2 California Skin Institute, Mountain View, CA 3 Department of Research & Evaluation, Kaiser Permanente Southern California; Kaiser Permanente Los Angeles Medical Center, Los Angeles, CA 4 Department of Pediatric Hematology/Oncology, Southern California Permanente Medical Group, Kaiser Permanente Los Angeles Medical Center, Los Angeles, CA 5 Department of Health Systems Science, Kaiser Permanente Bernard J. Tyson School of Medicine, Los Angeles, CA

Insurance status, a proxy for access to care, is an established correlate of cancer outcomes. Prior work in the field of healthcare disparities among melanoma patients, however, has included a mix of patients both with and without health insurance, making it difficult to disentangle the effects of various other sociodemographic factors. [1][2][3][4][5] In order to mitigate disparities and improve outcomes, we sought to independently examine the effects of these intertwined sociodemographic variables on all-cause mortality within an insured population of melanoma patients. Further, we aimed to evaluate the effects of health insurance coverage type, that is, whether patients were cared for within an integrated healthcare system or within a traditional model of healthcare, on all-cause mortality risk.
Our objective was to quantify the effect of race/ethnicity, socioeconomic status (SES) and healthcare system on overall mortality within an insured population of patients diagnosed with melanoma in Southern California from 2009 to 2014, and followed through 2017. Healthcare system was classified as those within Kaiser Permanente Southern California's (KPSC) network, a vertically integrated healthcare system, and insured patients outside of KPSC's network with other private insurance (OPI).
Using a retrospective cohort study design with data from the California Cancer Registry, we identified 14,614 adults diagnosed with melanoma (Stage 0-IV). The dataset included SES information based on geocoded data. The total number of deaths was 2,456 (16.8%) over a maximum follow up of 8 years. We examined person-year (PY) mortality rates and conducted Cox proportional hazard models, adjusted for age, sex, year of diagnosis, stage at diagnosis, race/ethnicity, SES, county of residence, and primary and adjuvant therapy.   Table 1 shows the distribution of demographics of the insured patients by healthcare system. KPSC had more minorities and those in the lowest two SES quintiles. Table 2 provides the PY all-cause mortality rates and multivariate adjusted hazard ratios (HR) by healthcare system. PY mortality rates by race/ethnicity did not yield significant results, possibly given the small numbers of deaths in certain populations. Mortality rates increased by decreasing SES quintile in the overall population. When stratifying by healthcare system, the PY mortality rates among those patients in KPSC were much more similar among SES groups, with the 95% confidence intervals (CI) overlapping for all five SES quintiles. By contrast, in OPI, the CIs for the lowest, lowermiddle, and middle SES groups did not overlap with the CIs of the upper-middle and highest SES groups. Of note, the KPSC patients in the three lowest SES quintiles had statistically significant decreased mortality rates compared to their OPI counterparts.
In multivariable adjusted hazard models, we did not observe differences in mortality risk by race/ethnicity in either healthcare system, when using Non-Hispanic Whites as the reference population. We did appreciate an increased mortality risk by decreasing SES quintile in KPSC and OPI, when using the highest SES quintile as the reference population. This trend, however, was much more apparent in the OPI group. For example, the poorest patients in OPI had a mortality risk 80% greater than wealthiest patients in OPI (HR 1.80; 95% CI 1.47, 2.22), while the poorest patients in KPSC had a 47% greater risk than the wealthiest patients in KPSC (HR 1.47; 95% CI 1.09, 2.00).
In summary, our results suggest that disparities in overall mortality persist, even in a cohort with health insurance coverage, and that lower SES is an important driver of this disparity. We also underscore the survival advantages for those vulnerable populations cared for within an integrated healthcare network, such as KPSC.