The effect of under and uninsured status on survival in lung cancer while adjusting for other mortality risk factors.
e21734 Background: Lung cancer is the leading cause of cancer death in the US. Significant improvements in survival have occurred with improved treatments. Payer status has been recognized as a barrier to treatment access across multiple cancer types including lung. This study aims to evaluate the impact of payer on 2-year survival rates for patients presenting in one health system with newly diagnosed lung or bronchus cancer. Methods: This is a retrospective survival analysis. There were 1,681 patients with lung cancer (small and non-small) under observation from time of first diagnosis. The first diagnosis date in the Cone health system defined study entry. Failure was defined as death during the 2-year observation period with right censoring after 2-years. Patients were categorized as underinsured if they had no insurance or Medicaid while those with commercial and Medicare were considered having full coverage. Cox proportional hazard models were used reporting hazard ratios. Results: Mortality rates per 10 patients diagnosed with lung or bronchus cancers were 3.5 for those with commercial insurance, 3.8 for Medicare, 3.3 for Medicaid and 5.4 for uninsured patients. Of those patients considered underinsured, 56.7% presented with stage IV cancer compared to those with full coverage (41.4%)*. 40.7% of those without insurance or underinsured were current tobacco product users compared to 25.1% of those with full coverage. Cox proportional hazard models revealed the risk of death is 1.34* times (95% CI 1.07-1.68) greater for underinsured patients compared to those with full coverage. The model adjusted for age, race, gender, marital status, language, Gini coefficient, Elixhauser comorbidity index, illicit drug use, cigarette smoker, smokeless tobacco user, alcohol use, PCP on record, and religion. However, when we add the AJCC stage to this model, the underinsured estimate is no longer statistically significant (1.23; 95% CI 0.92-1.65). Conclusions: Patients without insurance are diagnosed at later stages of disease. This late diagnosis is the primary driver of poor survival. Although being underinsured or uninsured is associated with a greater risk of death after diagnosis, adjusting for stage mitigates this effect. These findings support the need for equal access to early screening and diagnosis regardless of payer. *statistically significant at p < 0.05.