Abstract
Background/Aim: Melanoma incidence has increased in the United States over the past few decades, and disparities in patient treatment have been described. Although most patients with melanoma are good candidates for curative treatment, some are considered poor candidates for treatment because of comorbid conditions. We examined whether patient demographics influence treatment contraindication in melanoma. Patients and Methods: The National Cancer Database (NCDB) was used to identify patients with melanoma from 2004 through 2015. Multivariate logistic regression was used to determine independent associations, adjusted for confounders. We excluded patients who did not receive treatment for reasons and patients with unknown treatment status. Results: A total of 499,092 patients met the inclusion criteria. Of these, 525 (0.1%) had Treatment contraindicated because of comorbid conditions (TCBC) and 498,567 (99.9%) received treatment. Multivariate logistic regression showed higher odds of TCBC in patients with government insurance (OR=1.34, 95%CI=03-1.73; p=0.03) and patients without insurance (OR=2.75, 95%CI=1.76-4.29; p<0.001) than patients with private insurance. Conclusion: Demographic disparities affects treatment decision in oncological patients. Our study demonstrated a significantly higher likelihood of “nontreatment because of comorbid conditions” among melanoma patients with government insurance or without insurance. Greater efforts are needed to address inequalities in melanoma treatment in the United States.
Melanoma is an aggressive skin cancer and an important health issue worldwide (1). Melanoma incidence has increased in the United States over the past few decades, and disparities in patient treatment have been described (2-5). Surgical excision is essential for curative treatment of melanoma (6). Our aim was to analyze the characteristics associated with melanoma patients whose primary site surgery was not recommended or performed due to patient risk factors. We hypothesized that some patients may not receive needed treatment on the basis of modifiable factors.
Patients and Methods
This analysis was conducted with data from the National Cancer Database (NCDB), an initiative driven by the American Cancer Society and the American College of Surgeons’ Commission on Cancer. The NCDB is a hospital based dataset and registers 70% of all cancer diagnoses in the United States (7). This study was considered nonregulated by the Mayo Clinic Institutional Review Board.
Data were extracted for all patients with a diagnosis of melanoma from January 1, 2004, through December 31, 2015. The cohort was then split into 2 groups: 1) “Surgery of primary site was performed”; or 2) “Surgery of primary site not recommended/performed due to patient risk factors.” Patients were excluded if their surgery was not part of their planned treatment or if they refused surgery, died before surgery, or had unknown status regarding whether surgery was recommended or performed.
Demographic, facility, treatment, and tumor characteristics were collected. Patient demographic characteristics included age, sex, race/ethnicity, income (median household income of the patient’s ZIP code), education (percentage of adults who did not graduate from high school in the patient’s ZIP code), insurance status (uninsured or private or government insurance), and population density of the patient’s ZIP code. Facility characteristics included facility type. Tumor characteristics included invasive tumor behavior, Breslow depth, American Joint Committee on Cancer stage, and location on the body. Presence of comorbid conditions was evaluated with the Charlson/Deyo Comorbidity Score [0 (no comorbidity), 1, or 2 or more].
Demographic, facility, treatment, and tumor characteristics were analyzed with the χ2 or Mann-Whitney test, as appropriate. Multivariate analysis with logistic regression was performed to assess independent associations, adjusting for confounders. The outcome variable was “treatment contraindicated because of comorbid conditions (TCBC)” and the predicted variables were patient comorbid conditions (i.e., Charlson/Deyo Comorbidity Score) and demographic characteristics (age, race/ethnicity, sex, and income, education, and population density based on the patient’s ZIP code). Moreover, the logistic regression model was adjusted for potential tumor-related confounders (tumor stage, location on the body, and presence of metastasis). The significance level was set at p<0.05. Statistical analysis was done with SPSS statistical software version 25.0 (SPSS Inc).
Results
A total of 499,092 patients met the study criteria in the National Cancer Database (NCDB) of Melanoma. Of these, 525 (0.1%) had treatment contraindicated because of comorbid conditions (TCBC) and 498,567 (99.9%) received treatment. Interestingly, 71.8% of the patients with TCBC had a Charlson/Deyo Comorbidity Score of 0. Patients with TCBC were older than those who received treatment (69.63±15.51 years vs. 61.37±16.16 years). Most patients with TCBC were men (63.6%) and had government insurance (66.1%), invasive tumor (94.9%), stage IV cancer (60.8%), and metastasis at diagnosis (53.5%) (Table I).
Patient demographics by surgical treatment indication.
Multivariate logistic regression showed that older age and increased Charlson/Deyo Comorbidity Score were independently associated with higher odds of TCBC. However, we also observed higher odds of TCBC in patients with government insurance (OR=1.336, 95%CI=1.032-1.728; p=0.03) or without insurance (OR=2.751, 95%CI=1.764-4.290; p<0.001) than patients with private insurance. Patients with metastasis at diagnosis (OR=4.976, 95%CI=3.327-7.440; p<0.001), stage III cancer (OR=4.542, 95%CI=2.767-7.457; p<0.001), or stage IV cancer (OR=15.268, 95%CI=8.822-26.424; p<0.001) had higher odds of TCBC than patients with stage 0 cancer. Moreover, patients with tumors located in the trunk (OR=0.590, 95%CI=0.418-0.831; p=0.003) and extremities (OR=0.453, 95%CI=0.324-0.633; p<0.001) had lower odds of TCBC than patients with head and neck tumors (Table II).
Odds of having surgery contraindicated due to presence of risk factor.
Discussion
Patient demographics influence access to oncologic care (5, 9). We noted that 71.8% of patients with TCBC had a Charlson/Deyo Comorbidity Score of 0. Further analysis showed that type of insurance was an independent predictor of nontreatment. After adjustment for comorbid conditions and tumor severity, patients with government insurance or without insurance had higher odds of TCBC than patients with private insurance. Thus, our data suggest the need for efforts to address disparities in indications for melanoma treatment.
Authors have demonstrated the effect of health insurance on melanoma care (10, 11). Amini et al. (10) analyzed data from the Surveillance, Epidemiology, and End Results database to determine whether health insurance affected disease outcomes of 61,650 patients with melanoma. They noted that patients with Medicaid insurance or without insurance were more likely to have advanced disease at diagnosis and were less likely to receive curative treatment (10). In our analysis, patients with government insurance or without insurance had higher odds of TCBC than patients with private insurance.
Treatment delays have been associated with increased risks of comorbid conditions and death among cancer patients (12-15). Adamson et al. (11) demonstrated disparities in surgical treatment delays in a cohort of 7,629 patients with melanoma and included in the North Carolina Cancer Registry from 2004 through 2011. They noted that patients with private insurance experienced less delay in treatment than patients with Medicaid insurance (11).
Studies of national cancer databases are limited by their retrospective nature, missing data, and possibility of inaccurate data records. We were unable to account for histologic subtype and tumor mitotic index, limiting further analysis of tumor severity in patients with TCBC. Moreover, the NCDB only includes patients treated at hospitals. Nonetheless, the NCDB includes approximately 70% of the oncologic patients in the United States, thus providing this study with enough statistical power for multivariate logistic regression and to control for relevant confounding factors such as tumor characteristics. We encourage future investigations of TCBC in patients with melanoma, as well as how to promote clinical guidelines for fair administration of melanoma care, regardless of patient demographic characteristics.
Conclusion
Demographic disparities may influence the decision not to treat a patient because of comorbidities. Multivariate analysis, adjusted for confounders, showed that patients with government insurance or without insurance were more likely to have treatment contraindicated because of comorbid conditions. We hope that this finding supports future translational initiatives to reduce disparities in melanoma care.
Acknowledgements
This study was supported in part by the Plastic Surgery Foundation, Mayo Clinic Center for Individualized Medicine and Mayo Clinic Center for Regenerative Medicine.
Footnotes
Authors’ Contributions
DB, MTH, GG, FRA, and AJF had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: DB, AJF, SPB, ACS, XL, BDR. Acquisition, analysis, or interpretation of data: DB, MTH, FRA, GG, and AJF. Drafting of the manuscript: DB, MTH, FRA, GG, ACS. Critical revision of the manuscript for important intellectual content: SPB, ACS, XL, BDR, AJF. Study supervision: AJF.
Conflicts of Interest
The Authors have no conflicts of interest to declare regarding this study.
- Received February 7, 2021.
- Revision received February 24, 2021.
- Accepted March 3, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.