Changes in cancer mortality after Medicaid expansion and the role of stage at diagnosis.

BACKGROUND: Medicaid expansion is associated with improved survival following cancer diagnosis. However, little research has assessed how changes in cancer stage may mediate improved cancer mortality or how expansion may have decreased population-level cancer mortality rates. METHODS: Nationwide state-level cancer data from 2001 to 2019 for individuals ages 20-64 years were obtained from the combined Surveillance, Epidemiology, and End Results National Program of Cancer Registries (incidence) and the National Center for Health Statistics (mortality) databases. We estimated changes in distant stage cancer incidence and cancer mortality rates from pre- to post-2014 in expansion vs nonexpansion states using generalized estimating equations with robust standard errors. Mediation analyses were used to assess whether distant stage cancer incidence mediated changes in cancer mortality. RESULTS: There were 17 370 state-level observations. For all cancers combined, there were Medicaid expansion-associated decreases in distant stage cancer incidence (adjusted odds ratio = 0.967, 95% confidence interval = 0.943 to 0.992; P = .01) and cancer mortality (adjusted odds ratio = 0.965, 95% confidence interval = 0.936 to 0.995; P = .022). This translates to 2591 averted distant stage cancer diagnoses and 1616 averted cancer deaths in the Medicaid expansion states. Distant stage cancer incidence mediated 58.4% of expansion-associated changes in cancer mortality overall (P = .008). By cancer site subgroups, there were expansion-associated decreases in breast, cervix, and liver cancer mortality. CONCLUSIONS: Medicaid expansion was associated with decreased distant stage cancer incidence and cancer mortality. Approximately 60% of the expansion-associated changes in cancer mortality overall were mediated by distant stage diagnoses.