Breast cancer stage at diagnosis in relation to duration of Medicaid enrollment

BACKGROUND. Stage at diagnosis has been used to compare the quality of canc er screening services by health insurance type, using membership at diagnos is or treatment. This study evaluates breast cancer stage among women on Me di-Cal, California's Medicaid program, in relation to duration of coverage to assess the impact of including women with recently acquired benefits in the Medi-Cal group. METHODS. Breast cancers diagnosed in 1993 among women ages 30 to 64 were ob tained from the statewide, population-based cancer registry and linked to M edi-Cal enrollment files. Women on Medi-Cal when diagnosed were categorized based on months covered during the 12 months preceding diagnosis (12, 1-11 , or none), and compared with all other women with breast cancer. Logistic regression models measured the effect of duration of Medi-Cal coverage on t he odds of late-stage disease, controlling for demographic, socioeconomic, health access, and tumor characteristics. RESULTS. Among women with Medi-Cal benefits when diagnosed, 18% were not co vered during the year preceding diagnosis, and late-stage disease was commo n among these women. The odds ratio for late-stage disease among all women on Medi-Cal was 1.67 (95% CI 1.41, 1.97), but was reduced by 42% to 1.39 (9 5% CI 1.15,1.67) when women without benefits before diagnosis were excluded from the Medi-Cal group. CONCLUSIONS. Women with Medi-Cal benefits before diagnosis were more likely to be diagnosed with late-stage disease than other women with breast cance r. However, the practice of assigning health insurance status based on enro llment at diagnosis underestimates the effect of access to breast cancer sc reening through Medicaid.