In-Hospital Cancer Mortality Risk: Variation by Rural-Urban Setting and Payer

Persons living in rural areas may experience difficulties in accessing timely and high quality care for cancer treatment, although research on this question is limited. In addition, persons who are uninsured or covered by Medicaid may enjoy poorer cancer treatment access and quality compared to persons with private insurance. The present study used the 2009 State Inpatient Database (part of the Healthcare Cost and Utilization Project) to investigate variation in risk of mortality during hospitalization for patients according to rural-urban setting and payer. Rural setting was measured by rural-urban commuting codes and classified as whether the patient resided in a city, large rural town, or small rural town/isolated area. Primary payer was measured as Medicaid, private pay, self-pay (uninsured), or other payer; Medicare patients were excluded so that the focus on this paper was on the population under age 65. Patients with a primary cancer diagnosis were included (N=143,648 with complete data on measures of interest). Overall, 4.46% of cancer patients died during the hospital stay. Before adjusting for covariates, mortality rates were significantly higher in rural areas versus urban areas, and in patients with Medicaid or patients without insurance relative to private pay patients. In logistic regression analysis that adjusted for age, race/ethnicity categories, sex, and co-morbidities, patients without insurance remained at significantly elevated mortality risk (OR=1.62, 95% CI=1.30, 2.03). In addition, although there were not significant main effects for rural setting or Medicaid in the adjusted model, there was a significant interaction whereby patients with Medicaid living in rural areas were at greater risk for in-hospital mortality (OR=1.13, 95% CI=1.02, 1.27). In summary, the results suggest that among persons under age 65, in-hospital cancer mortality disparities are greater among rural patients and patients without private insurance, and that rural disparities are partially accounted for by poorer outcomes among Medicaid patients. Poorer Medicaid outcomes in rural areas might reflect rural Medicaid disadvantages in pre-admission cancer treatment, disadvantages in rural-based hospital cancer care, or other unmeasured characteristics of rural Medicaid patients. The results highlight the need to provide equitable cancer prevention and treatment across populations characterized by geography and payer.