Medicaid expansion and trends in cancer stage at diagnosis.

Abstract

e14129 Background: Cancer stage distribution offers an early window into the effect of the ACA’s 2014 Medicaid expansion. If Medicaid expansion enabled uninsured adults to access preventive care, then the proportion of cancers diagnosed at early stage should be higher in states that adopted it. Methods: Patient level data were retrieved from SEER-18, New York and Texas cancer registries from January 2010 to December 2015, for eight common tumor types; including N = 2,493,589 (breast, cervical, colorectal, prostate, lung) screen detectable and N = 369,203 (testis, kidney and uterine corpus) not detected by routine screening. Patients were categorized by residence in states that were: 1) Non-expanders; 2) Expanders - income eligibility thresholds increased from 0% to 138% of the federal poverty level (FPL) in 2014; and 3) Pre-Expanders, with pre-2014 eligibility thresholds > 70% of FPL. Stage distributions were compared before and after January 2014, categorized by AJCC stage and using ordinal stage equivalent (OSE) which assigns a numeric score based on AJCC stage. Difference in difference in difference (D3) methods compared patients ≥65 versus < 65, and screen-detectable versus not. Multivariable logistic regression was used to adjust for variation in age, sex, race, ethnicity and tumor site across states. Results: No differences in stage distribution were observed across the three categories of state. (Table). D3 analysis revealed no difference between trends in the over 65s, or for screen-detectable cancers. Adjustment for age, sex, race, ethnicity and tumor site made no significant difference to these results. Conclusions: States’ Medicaid expansion status in 2014 was not associated with a significant difference in early stage diagnoses for adults with cancer diagnosed before age 65. Obstacles to early cancer detection other than financial access must be addressed to improve population-level outcomes. [Table: see text]