Introduction:

Adolescent and young adult (AYAs: 15-39) patients with cancer have not had the same relative improvement in survival as other age groups over the last decades. Studies have shown that having public insurance or being uninsured at diagnosis is associated with more advanced disease at presentation and worse overall survival. However, previous studies have not differentiated patients who joined Medicaid at diagnosis from those with continuous enrollment which may have different implications for access to care prior to diagnosis. Therefore, we examined the impact of insurance status, including Medicaid enrollment at diagnosis, on stage at diagnosis for AYAs with non-Hodgkin lymphoma (NHL) and Hodgkin lymphoma (HL) only] and on survival for AYAs with NHL, HL, acute myeloblastic leukemia (AML), and acute lymphoblastic leukemia (ALL).

Methods:

Using Medicaid enrollment data linked to the California Cancer Registry, we identified AYAs with NHL, HL, ALL, and AML diagnosed from 2005 to 2014. Insurance type was classified as: continuous Medicaid, discontinuous Medicaid prior to diagnosis, Medicaid at diagnosis, other public (Medicare, Indian/Public Health Service, county), private/military, and uninsured. Multivariable logistic regression and Cox proportional hazards regression were used to determine the impact of insurance type on stage at diagnosis (for NHL and HL) and overall survival, respectively. Results are represented as adjusted odds ratios (OR) and hazard ratios (HR) with associated 95% confidence intervals (CI).

Results:

Of the 11,667 AYA patients in our study, 4,435 had NHL, 4,161 had HL, 1,522 had AML and 1,549 had ALL. Patients with HL had the highest proportion of private insurance (66%) followed by those with NHL (60%), AML (50%) and ALL (37%). Of the 4,059 patients enrolled in Medicaid, 41% had continuous Medicaid, 15% had discontinuous Medicaid and 43% received Medicaid at diagnosis. Only 2-4% of patients, depending on primary diagnosis, remained uninsured after cancer diagnosis. The majority of AYAs with HL and NHL were diagnosed with stage I/II disease (59% and 52% respectively). Compared to AYAs with private insurance, NHL and HL patients with discontinuous Medicaid and Medicaid at diagnosis had a higher likelihood of later stage disease (III-IV vs I/II) at diagnosis (NHL: discontinuous OR 1.45, CI 1.10-1.92; at diagnosis OR 1.69, CI 1.38-2.06; HL: discontinuous OR 1.63, CI 1.19-2.23; at diagnosis OR 1.68, CI 1.35-2.09) after adjusting for sociodemographic factors, baseline comorbidities and type of facility. In addition, NHL patients with continuous Medicaid (OR 1.23, CI 1.01, 1.51) and HL patients with other public insurance (OR 1.56, CI 1.05-2.32) had a higher odds of late stage disease. Type of health insurance was associated with overall survival in multivariable models (Table). NHL patients with Medicaid (continuous HR 1.74, CI 1.39-2.17; discontinuous HR 2.52, CI 1.94-3.27; at diagnosis HR 1.88, CI 1.53-2.31), other public (HR 1.83, CI 1.16-2.87) and no insurance (HR 1.87, CI 1.09-3.20) had worse survival than NHL patients with private insurance. Similarly, HL patients with Medicaid (continuous HR 2.10, CI 1.42-3.12; discontinuous HR 1.89, CI 1.08-3.29; at diagnosis HR 2.43, CI 1.699-3.48) and no insurance (HR 1.87, CI 1.09-3.20) experienced worse survival. For AML, health insurance was not significantly associated with survival. For ALL, only continuous Medicaid (HR 1.32, CI 1.05-1.67) and other public (HR 1.32, CI 1.05-1.67) insurance were associated with worse survival, though discontinuous Medicaid trended toward significance (p=0.06).

Conclusion:

Our study demonstrates that a significant proportion of patients previously thought to have public insurance were discontinuously insured with Medicaid or uninsured at time of diagnosis, only receiving Medicaid after diagnosis. While important, insurance enrollment at diagnosis does not provide the same pre-diagnosis access to services as those with continuous enrollment. Indeed, for NHL and HL, we observed the strongest associations between discontinuous Medicaid and Medicaid at diagnosis and late stage disease. However, Medicaid, regardless of type of enrollment, was associated with worse survival in AYAs with NHL, HL and ALL relative to private insurance. Therefore, future studies should focus on factors influencing worse outcomes for AYA patients with public insurance.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution