Association between Medicaid-enrollment status/timing and stage at diagnosis

Objective: Prior studies have shown that both uninsured and Medicaid-insured cancer patients have an increased risk of advanced stage at diagnosis compared to Medicare- or privately-insured patients. At the national level, such studies have not been able to evaluate differences in stage among patients who were uninsured at the time of diagnosis and subsequently enrolled in Medicaid and those enrolled in Medicaid prior to diagnosis. However, studies in two states (Michigan and California) have reported later stage at diagnosis among patients enrolled in Medicaid after (versus before) diagnosis. The current study will assess stage at diagnosis among cancer patients in South Carolina by Medicaid-enrollment status and timing for 12 cancer sites. We hypothesize that patients who enrolled in Medicaid after cancer diagnosis will have more advanced stage disease than those who are enrolled prior to diagnosis. Methods: Retrospective analyses were conducted of secondary data from the South Carolina Cancer Registry and Medicaid program. Patients ages 18-64 diagnosed during 1996-2005 with a first invasive primary cancer of the female breast, lung/bronchus, prostate, colon, rectum, kidney/renal pelvis, thyroid, pancreas, ovary, or urinary bladder and patients with melanoma (skin) or non-Hodgkin's lymphoma with non-missing vital status or cancer stage data were selected for analysis. Multivariate logistic regression analysis was used to examine the relationship between Medicaid-enrollment status (non-Medicaid as referent) and stage at diagnosis controlling for all significant (p< .05) predictors. Early stage was defined as localized disease. Late stage was defined as regional/distant disease. Patients continuously enrolled in Medicaid for at least the month prior to and the month of diagnosis were considered enrolled prior to diagnosis. Patients enrolled in the month of diagnosis or thereafter and not Medicaid-enrolled in the month prior to diagnosis were considered late enrolled. Results: Of 48,099 evaluable patients, late Medicaid enrollees were more likely (p < .00001) to be late stage (79%) than patients enrolled prior to diagnosis (59%) or non-Medicaid-enrolled patients (43%). Compared to non-Medicaid patients, the odds of being late stage were 2.88 (C.I.=2.61-3.19) for late Medicaid enrollees and 1.25 (C.I.=1.15-1.35) for patients enrolled prior to diagnosis after controlling for all other factors. Conclusion: Differences in stage between Medicaid-insured and non-Medicaid-insured patients were generally greatest for those cancer sites for which screening is recommended. For almost all cancer sites, patients who were Medicaid enrolled after their cancer diagnosis were significantly more likely to be diagnosed with late stage disease.