Association of county-level medical debt and survival among individuals newly diagnosed with cancer in the United States.

156 Background: Individuals diagnosed with cancer often bear significant health care costs and productivity loss, with over half incurring cancer-related medical debt. Medical debt is increasingly recognized as a social determinant of health care and health outcomes. Using newly available nationwide county-level medical debt information from 2011 to 2019 linked with national cancer registry data, we examined the association of county-level medical debt with survival among individuals newly diagnosed with cancer in the US. Methods: Individuals aged ≥18 years with newly diagnosed first primary cancer in 2011-2019 were identified from the National Cancer Database, which was linked with credit bureau data on county-level percentage of adults with medical debt in collections based on patients’ year of diagnosis. Patients were followed from diagnosis through December 31, 2019. Hazard Ratios (HR) and were derived from Cox proportional hazard models to estimate the associations of county-level medical debt quartiles with overall survival, adjusting for year of diagnosis, age, sex, race and ethnicity, insurance status, comorbidity score, metropolitan status, county-level social deprivation index and medically underserved area designation, and state of residency. Stratified analyses were conducted by cancer type and sociodemographic and clinical factors. Results: A total of 7,558,658 individuals newly diagnosed with cancer were identified, with a median county-level medical debt of 18%, ranging from 0% to 56%. County-level medical debt decreased since 2015. The highest percentage of county-level medical debts was found among patients racialized as non-Hispanic Black, uninsured, and living in medically underserved areas. In adjusted analyses, patients living in counties with the highest quartile of medical debt had 17% increased risks of death compared with those living in counties with the lowest quartile of medical debt (HR=1.17, 95%CI=1.16-1.18), with a statistically significant dose-response relationship for the higher quartiles (P-trend<.0001). The pattern was consistently seen across cancer types as well as socioeconomic and clinical groups. Conclusions: County-level medical debt was strongly associated with worse survival among individuals newly diagnosed with cancer, consistent across socioeconomics and clinical subgroups. Policies preventing and relieving medical debt maybe effective ways to improve cancer outcomes at population level. Hazard ratios for risk of death associated with county-level medical debt by cancer site and stage at diagnosis. County-Level Share of Population with Medical Debt Q1 Q2 Q3 Q4 All cancers 1 1.08 1.13 1.17 Female breast 1 1.10 1.13 1.16 Prostate 1 1.06 1.11 1.13 Lung 1 1.06 1.10 1.12 Colon and rectum 1 1.07 1.11 1.14 Stage I-III 1 1.09 1.14 1.19 Stage IV 1 1.07 1.11 1.13 All P for trend <.0001.