Association between financial fragility and treatment patterns in multiple myeloma.

Abstract

11085 Background: Multiple myeloma (MM) drugs are associated with significant out-of-pocket costs. We hypothesize that patients with financial fragility (FF) may receive suboptimal MM treatment, as they might be less likely to be able to access and afford these medications compared to patients in good financial standing. We examined the association between FF and MM treatment by using a novel database that links patient-level credit records to cancer and claims data in Washington (WA) state. Methods: We conducted a retrospective analysis of newly diagnosed patients with MM (2012-2020) with Medicare and commercial health insurance, via a database linking WA cancer registry data (including Western WA SEER), health insurance claims, and depersonalized credit reports (TransUnion). All eligible patients received combination therapy with an oral immunomodulatory drug and at least one injectable medication within 6 months of diagnosis. FF was defined as evidence of at least one of the following in credit reporting around diagnosis (50 days before or after): charge-offs, collections, liens, foreclosures, repossessions, and bankruptcy in the last 12 months. Patients were categorized as receiving suboptimal treatment (the composite outcome) if they experienced at least one of the following: delay in time to treatment initiation (>40 days from diagnosis), treatment interruption (gap of >30 days without treatment in the first 6 months), or lack of autologous stem cell transplantation (ASCT, within the first 12 months). We performed chi-squared tests followed by multivariable logistic regressions, adjusting for age, sex, race (white/nonwhite), insurance type, and Area Deprivation Index (dichotomized; 1-5 and 6-10). Results: A total of 204 eligible patients (median age 69 years, 60% male, 91% White, and 60% Medicare-insured) met eligibility criteria. Of these, 39 patients (19%) had evidence of FF and 120 patients (59%) met the composite outcome. FF was significantly associated with the composite outcome of suboptimal treatment in bivariate analysis (74% vs. 55%, p=0.03) and adjusted multivariable analysis (OR 2.41 [95% CI: 1.05-5.51], p=0.04). As 182 (89%) patients received ASCT within 12 months in this cohort, we performed a sensitivity analysis by removing ASCT as a component outcome; in this analysis, 108 patients (53%) met the composite outcome. FF was associated with the composite outcome of suboptimal treatment in bivariate analysis (69% vs. 49%, p=0.02) and adjusted multivariable analysis (OR 2.31 [95% CI: 1.05-5.10], p=0.04). Conclusions: FF is significantly associated with suboptimal MM treatment patterns, including delay to treatment initiation, treatment interruption, and decreased receipt of ASCT. Larger prospective studies with more diverse demographic representation and incorporation of clinical data are in development to elucidate the mechanisms through which FF leads to suboptimal MM treatment.