e23036 Background: Prior population-based registry studies found Medicaid expansion is associated with improved cancer survival, but registries often lack comorbidity and treatment-setting detail. The National Cancer Database (NCDB) includes these factors, enabling assessment of expansion-associated survival effects in hospital-based data. We used a difference-in-differences (DiD) approach to evaluate expansion-associated changes in Medicaid-at-diagnosis and 5-year overall survival (OS) across five cancers spanning a range of baseline survival. Methods: We included 1,400,501 adults from NCDB aged 18–64 at time of diagnosis with breast, cervical, colon, non-small cell lung (NSCLC), or pancreatic cancer from 2005-16. States were dichotomized as Medicaid expansion in 2014 (AR, AZ, CO, DE, HI, IA, IL, KY, MA, MD, ND, NM, NV, NY, OH, OR, RI, VT, WV) vs non-expansion (AL, FL, GA, ID, KS, ME, MO, MS, NC, NE, OK, SC, SD, TN, TX, UT, VA, WI, WY). Seven late expansion states were omitted. Using DiD, we quantified (1) changes in proportion covered by Medicaid and (2) estimated 5- year OS using multivariable Cox models adjusted for comorbidity, treatment-setting factors and robust standard errors clustered by facility.Primary analyses did not include area-level income/education quartiles and urbanicity; sensitivity analyses added these covariates. Pre- expansion trends for Medicaid-at-diagnosis and 5-year OS were similar across cancers, except pancreatic cancer. Results: Medicaid-at-diagnosis increased more in expansion states across cancers (DiD +3.9 to +10.8 percentage points). In DiD Cox models of 5-year OS, expansion was significantly associated with lower mortality for colon cancer (HR 0.68–0.69) and NSCLC (HR 0.75) across primary and sensitivity analyses; cervical (primary HR 0.84, 95% CI 0.64–1.10) and pancreatic cancer (primary HR 0.83, 0.66–1.04) were not significant in either analysis (Table). Breast cancer was borderline in sensitivity (HR 0.68, 0.46–1.00) but significant in the primary analysis (HR 0.67, 0.47–0.96). Conclusions: In a multi-cancer, quasi-experimental DiD analysis of the NCDB, Medicaid expansion was associated with a higher proportion of patients with 5 cancer types on Medicaid and improved 5- year OS for patients with colon cancer and NSCLC, with breast cancer showing significant association only in primary analyses; cervical and pancreatic cancer showed no significant differences. These findings add to evidence that Medicaid expansion can improve cancer survival, with benefits that vary by cancer type. DiD estimates for 5-year OS (Cox models). Cancer N Primary aHR (95% CI) Sensitivity Analysis aHR (95% CI) Breast 816,418 0.67 (0.47-0.96) 0.68 (0.46-1.00) Cervical 40,684 0.84 (0.64-1.10) 0.80 (0.59-1.08) Colon 176,529 0.68 (0.52-0.88) 0.69 (0.53-0.91) NSCLC 292,401 0.75 (0.59-0.96) 0.75 (0.58-0.97) Pancreas 74,469 0.83 (0.66-1.04) 0.82 (0.64-1.05)
Copeland et al. (Thu,) studied this question.