Relationship between insurance status and diagnosis of cutaneous immune-related adverse events.

Abstract

e18535 Background: Cutaneous immune-related adverse events (cirAEs) are among the most common side effects of immune checkpoint inhibitor (ICI) therapy. While insurance status has been shown to influence outcomes in patients treated with ICIs, its impact on cirAE management remains underexplored. We therefore evaluated insurance status in patients with cirAEs, examining its effect on rate of and time to cirAE diagnosis. Methods: Using billing data, we retrospectively identified patients who initiated anti-PD-1/PDL-1 and/or anti-CTLA-4 therapy at Massachusetts General Hospital between January 1, 2016 and March 8, 2019 (n = 2,459) for possible cirAE. Eligible cirAEs included reactions attributed to ICI by the clinician, consistent with established morphologic categories. For each patient with confirmed cirAE (n = 358), we abstracted oncologic history, cirAE features, and insurance status. Associations between insurance and cirAE diagnosis outcomes were assessed via logistic and linear regression, and adjusted for age, sex, race, ICI type, cancer diagnosis, cirAE type, and significant covariates ( P< 0.05). Results: Of the 2,459 patients who received ICIs, 2,419 (98.4%) had documented insurance status. Most ICI recipients had Medicare (n = 1,119; 46.3%) or private insurance (n = 1,156; 47.8%) relative to Medicaid (n = 104; 4.3%) or other government insurance (e.g. Tricare) (n = 40; 1.7%). We found that 358 (median age 64 years, 40.5% female) developed a cirAE. Among cirAE patients, 175 had Medicare (48.9%), 174 had private insurance (48.6%), 6 had Medicaid (1.7%), and 3 had other government insurance (0.8%). The most common cirAEs across insurance types were maculopapular rash, pruritus, and eczematous and lichenoid eruptions. In the multivariable analysis, ICI patients with Medicare insurance had a higher rate of cirAE diagnosis (adjusted odds ratio: 2.41, 95% CI: 1.00, 5.90, P= 0.05) relative to Medicaid patients. In addition, in terms of time to cirAE diagnosis at dermatology visit, Medicare insurance was associated with longer delays, with a linear regression coefficient of 132.2 (95% CI: 4.78, 259.6; P= 0.04). No significant associations were found between other insurance types and cirAE diagnosis outcomes. Conclusions: Our study shows that patients with Medicaid are less likely to be diagnosed with cirAE relative to those with Medicare, despite delays in diagnosis, when controlling for all other demographic/oncologic factors. Ultimately, these findings are reassuring that despite insurance differences, patients with cirAEs are receiving suitable care and appropriately seen by dermatologists. As insurance coverage for specialists can vary widely, these initial findings are a promising indicator that patients with cirAEs are well-connected within healthcare systems.