Abstract C54: Survival outcomes for head and neck patients with Medicaid: A health insurance paradox

Abstract

Abstract Introduction: Although there are currently more than 430,000 head and neck cancer (HNC) survivors in the United States, it is accepted that many more patients would have survived longer if they presented at an earlier stage. Less than half of all head and neck cancer patients present with early-stage disease. One of the factors implicated in late stage of presentation for head and neck cancer patients is access to care, driven by health insurance status. While individuals with health insurance are known to present earlier, less is known about outcome differences for patients who are uninsured or who have Medicaid insurance. We have observed many head and neck cancer patients initially present without insurance despite qualifying for Medicaid, and so are assisted with obtaining insurance before discharge. This process blurs the line between uninsured and Medicaid patients. The aim of this study was to determine whether there are disparities in survival outcomes for HNC patients based on whether they are insured, uninsured, or have Medicaid insurance. Methods: A cohort of 49,524 patients aged 18-64 years with first primary HNC from the Surveillance, Epidemiology, and End Results (SEER) 18 database diagnosed from 2007-2014 was included. Actuarial survival curves stratified by insurance status (insured, Medicaid, and uninsured) were created to determine HNC-specific survival differences between the groups with a log-rank test. Patient characteristics including insurance, race/ethnicity, sex, county-level poverty, surgery, marital status, tumor site, stage, year of diagnosis, and age at diagnosis were utilized in a Fine and Gray competing risk proportional hazard model to compute adjusted hazard ratios (aHR) for cause-specific death from HNC. Multinomial logistic regression was also performed to determine characteristics of patients with each type of insurance by adjusted odds ratios (aOR). Results: The cohort was mostly male (75.6%) and insured (73.6%), with 18.6% on Medicaid and 7.8% uninsured. At the end of the 7-year follow-up period, HNC-specific survival rate was significantly lower for patients on Medicaid (49.5%) than uninsured (54.8%) and insured patients (74.2%) (log-rank p < 0.001). After controlling for other prognostic demographic and clinical factors, patients on Medicaid (aHR=1.63, 95% CI: 1.54, 1.71) and uninsured patients (aHR=1.33, 95% CI: 1.43, 1.64) had higher hazard of death from HNC compared to insured patients. Compared with non-Hispanic whites, other race/ethnicity groups were more likely to be on Medicaid, with non-Hispanic American Indians/Alaska Natives being the most likely (aOR=3.46, 95% CI: 2.66, 4.51). Medicaid were more likely not have received surgery than insured patients (aOR=1.34, 95% CI: 1.26, 1.42). They were also more likely to present at advanced disease. Conclusion: While patients with health insurance had better survival outcome in general, our study showed that patients with Medicaid did not have a better survival outcome than those without any insurance after adjusting for all other prognostic factors, including stage of presentation and treatment modality. Medicaid patients, in fact, had worse outcome than uninsured HNC patients. It could be that despite having insurance, Medicaid patients did not have adequate access to care and thus had delayed presentations. Our findings highlight the need to bridge the health insurance gap for HNC patients to increase survivorship. Citation Format: Nosayaba Osazuwa-Peters, Matthew C. Simpson, Sean T. Massa, Eric Adjei Boakye, Lauren M. Cass, Sai Deepika Challapalli, Rebecca L. Rohde, Mark A. Varvares. Survival outcomes for head and neck patients with Medicaid: A health insurance paradox [abstract]. In: Proceedings of the Tenth AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2017 Sep 25-28; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2018;27(7 Suppl):Abstract nr C54.