Ethnic disparities in melanoma diagnosis and survival The effects of socioeconomic status and health insurance coverage Amy M. Anderson-Mellies, MPH 1 Matthew J. Rioth, MD 1,2,3 Myles G. Cockburn, PhD 1,4,5,6 1University of Colorado Comprehensive Cancer Center, University of Colorado Anschutz Medical Campus, Aurora, CO; 2 Division of Biomedical Informatics and Personalized Medicine, School of Medicine, University of Colorado Anschutz Medical Campus, Aurora, CO; 3 Division of Medical Oncology, School of Medicine, University of Colorado Anschutz Medical Campus, Aurora, CO; 4 Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, CO; 5 Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA; 6 Department of Dermatology, Keck School of Medicine, University of Southern California, Los Angeles, CA
The big picture Stage Distribution for Melanomas by Ethnicity, 2008 to 2012 100 Five-Year Melanoma-Specific Survival Rates (%) by Ethnicity, 2005 to 2011 100 75 72 84 93 93 88 88 Percent 50 Percent 85 79 25 0 15 9 7 4 5 4 Localized Regional Distant Unknown 78 70 Male Female Hispanic NHW Hispanic NHW Data source: Cancer statistics for Hispanics/Latinos, 2015, Volume: 65, Issue: 6, Pages: 457-480, First published: 16 September 2015, DOI: (10.3322/caac.21314)
Some possible explanations Advanced stage Screening Thicker tumors Biology Hispanic Nodular & Acral Lentiginous melanomas Poorer prognosis Low SES Medicaid & Uninsured Access to Care
The Big Question Are Hispanic disparities in melanoma diagnosis and survival explained by modifiable factors rather than by tumor characteristics or other unmeasured biological factors?
Methods Data source: California Cancer Registry Study population: Non-Hispanic White (NHW) and Hispanic patients diagnosed with invasive cutaneous melanoma between 1995 and 2014 Variables: Age, sex, race/ethnicity, neighborhood socioeconomic status (SES), insurance status, SEER summary stage, histology, and anatomic site Outcomes: Late (regional or distant) stage at diagnosis and melanoma-related death (survival)
Tumor Characteristics Male Female NHW Hispanic NHW N=70,895 N=2,828 N=46,081 Hispanic N=3,729 Stage* Localized 83.0% 65.6% 86.9% 78.9% Regional 8.2% 17.0% 6.6% 10.9% Distant 4.7% 10.8% 3.0% 5.2% Unstaged 4.0% 6.6% 3.6% 4.9% Histology* Malignant melanoma, NOS 54.5% 53.9% 54.7% 52.4% Superficial spreading melanoma 25.8% 20.2% 29.9% 28.8% Nodular melanoma 7.3% 11.0% 5.9% 7.5% Lentigo maligna melanoma 5.7% 3.6% 3.9% 2.3% Acral lentiginous melanoma 0.5% 5.6% 0.8% 4.3% Other 6.2% 5.7% 4.8% 4.7% Anatomic Site* Limbs 31.4% 39.8% 58.0% 57.6% Head/Neck 26.3% 23.2% 14.5% 15.4% Trunk 37.5% 28.3% 24.1% 22.1% Other 4.8% 8.7% 3.4% 4.9% *p<.001
SES and insurance Male Female NHW N=70,895 Hispanic N=2,828 NHW N=46,081 Hispanic N=3,729 Age at Diagnosis (mean)* 63.6 yrs 58.1 yrs 58.4 yrs 52.1 yrs Socioeconomic Status* Highest SES 34.6% 14.8% 32.9% 15.4% Higher-middle SES 24.9% 16.8% 25.6% 20.3% Middle SES 18.3% 19.5% 19.4% 20.7% Lower-middle SES 12.0% 22.2% 12.6% 19.8% Lowest SES 5.7% 22.8% 5.7% 20.4% Unknown 4.5% 3.9% 3.8% 3.4% Insurance Status* Private 48.3% 43.7% 54.3% 51.9% Medicaid 2.0% 11.6% 2.4% 9.9% Medicare 28.4% 22.5% 22.7% 15.2% Military 3.1% 1.7% 0.9% 0.6% Insured, NOS 5.8% 4.6% 6.7% 5.5% Uninsured 1.5% 5.0% 1.4% 3.8% Unknown 10.8% 10.9% 11.6% 13.2% *p<.001
Late-stage disparities (males) Male, Hispanic Male, NHW Socioeconomic Status Insurance Status Histology 0 12.5 25 37.5 50 Percent diagnosed at regional/distant stage * ALM (acral lentiginous melanoma), LMM (lentigo maligna melanoma), NM (nodular), SSM (superficial spreading), MM(NOS) malignant melanoma not otherwise specified
Late-stage disparities (females) Female, Hispanic Female, NHW Socioeconomic Status Insurance Status Histology 0 15 30 45 60 Percent diagnosed at regional/distant stage * ALM (acral lentiginous melanoma), LMM (lentigo maligna melanoma), NM (nodular), SSM (superficial spreading), MM(NOS) malignant melanoma not otherwise specified
Putting it all in one model Hispanic Higher-middle SES Middle SES Lower-middle SES Lowest SES Medicaid Uninsured SSM NM LMM ALM Other * Referent groups are NHW, 0.0 MM(NOS), private insurance, 4.0and highest SES. Model 8.0 also adjusts for age, sex, and 12.0 anatomic site. 16.0 Adjusted* Odds Ratio
So There s something about being Hispanic (over and above tumor characteristics, lower SES and insurance coverage) that increases risk of being diagnosed with late stage disease
10 year 82.5% (82.1-82.8) 71.3% (69.0-73.4) 89.0% (88.6-89.3) 85.7% (84.3-87.0) What about survival? Survival Male, NHW Male, Hispanic Female, NHW Female, Hispanic 1 year 95.8% (95.7-96.0) 92.2% (91.0-93.2) 97.3% (97.1-97.4) 96.4% (95.7-97.0) 5 year 86.7% (86.4-86.9) 77.6% (75.8-79.4) 91.7% (91.4-92.0) 88.3% (87.1-89.4)
What explains survival differences? Model Hazard Ratio Unadjusted 1.39 (1.29-1.49) Adjusted for age and sex 1.72 (1.60-1.86) Adjusted for above and: Stage Histologic type Anatomic site Adjusted for above and: SES Insurance status 1.14 (1.06-1.23) 0.93 (0.86-1.01)
Probability of death Hispanic Higher-middle SES Middle SES Lower-middle SES Lowest SES Medicaid Uninsured Regional Distant SSM NM LMM ALM Other 0.0 4.5 9.0 13.5 18.0 *Referent groups are NHW, localized stage, MM(NOS), private insurance, and highest SES. Model also adjusts for age, sex, and anatomic site. Adjusted* Hazard Ratio
Summary Hispanics with melanoma are more likely to be diagnosed at a late stage even after accounting for SES, insurance status, histology, and anatomic site. Hispanics have an overall 72% increased risk of death from melanoma, and a 14% increased risk after considering stage at diagnosis and tumor characteristics However, accounting for SES and insurance status explains the remaining increased risk of death observed among Hispanics.
Conclusions 1. Later stage at diagnosis for Hispanics could be due to knowledge of risk and attitudes towards screening (not able to be explained by tumor characteristics nor SES/insurance) Plenty of evidence that Hispanics underestimate their risk of skin cancer, and don t think screening is important. Late stage disease results in substantial morbidity, long term health effects, loss of job etc, especially problematic among lower SES individuals Results in (avoidable) system-wide health care expenses, especially for Medicaid/Medicare
Conclusions 2. Poor Hispanic melanoma survival is attributable to: Presenting with melanoma at a late stage (so we need to fix #1!!) Insurance status and SES rather than ethnicity or tumor characteristics (which means they are preventable) o Ability to get to treatment? o Ability to stick with treatment? o Transportation and other barriers?
Conclusions 3. Improvements in Hispanic melanoma survival could be achieved by focusing on diagnosing Hispanic patients earlier. 4. More work is required to determine what barriers there are to melanoma diagnosis for Hispanics, and to determine if those barriers are also relevant to obtaining the appropriate follow-up care after diagnosis (survival).
Acknowledgements This work was supported in part by the Population Health Shared Resource of the University of Colorado Comprehensive Cancer Center funded by the National Cancer Institute grant P30CA046934. The collection of cancer incidence data used in this study was supported by the California Department of Public Health pursuant to California Health and Safety Code Section 103885; Centers for Disease Control and Prevention s (CDC) National Program of Cancer Registries, under cooperative agreement 5NU58DP003862-04/DP003862; the National Cancer Institute s Surveillance, Epidemiology and End Results Program under contract HHSN261201000140C awarded to the Cancer Prevention Institute of California, contract HHSN261201000035C awarded to the University of Southern California, and contract HHSN261201000034C awarded to the Public Health Institute. The ideas and opinions expressed herein are those of the author(s) and do not necessarily reflect the opinions of the State of California, Department of Public Health, the National Cancer Institute, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors.
Questions? amy.mellies@ucdenver.edu