Poster #23 - Harman Sabharwal

Integrative Visual Analytics of Sociodemographic and Financial Literacy Data to Identify High-Risk Veteran Populations

Ramona Raya, MD, Department of Medicine, University of Virginia School of Medicine; Harman Sabharwal, Langley High School, McLean, VA

Background: Veteran health insurance programs aim to reduce financial toxicity-the destabilizing impact of healthcare costs-yet housing and financial instability continue to rise. Using a bioinformatics-inspired stratification approach, this study analyzed sociodemographic variables and financial literacy data to identify high-risk veteran subgroups beyond cancer populations, where financial literacy research has traditionally focused. Methods: Adapted versions of the National Financial Capability Study and the Comprehensive Score for Financial Toxicity were used to survey n=88 United States military veterans. Descriptive, nonparametric (Chi-square and Fisher's exact tests), and logistic regression analyses were conducted. Financial literacy subdomains-retirement, inflation, and compound interest-were evaluated, with race and gender examined as interacting factors shaping financial toxicity. Results: Black and female veterans had significantly lower financial literacy scores compared to their White male counterparts (B = -0.49, p = 0.039 and B = -0.92, p = 0.016, respectively). Black veterans scored lower on retirement and inflation knowledge than non-Black peers (p=0.019, p=0.043), and women underperformed men on compound interest and inflation (p=0.041, p=0.026). Black Female veterans experienced higher financial toxicity: borrowing or using savings for healthcare (60%, 95% CI: 23.1%-88.2%, p < 0.001), lower full-time employment (17%, CI: 3.0%-56.4%), and greater health-related work reduction (80%, CI: 37.6%-96.4%, p = 0.015), compared to other race and gender groups. Use of veteran financial support programs was low among all veterans (26.9%) and not significantly higher among Black female veterans (p = 0.14) compared to all other veterans in the sample. Conclusion: This bioinformatics-style visual analysis showed that Black female veterans experience significantly greater financial toxicity than White and male counterparts, underscoring the need to apply machine learning to larger datasets to enable real-time, tailored outreach and improve financial resilience among high-risk veterans.