2913-LB: Social Vulnerability Moderates the Protective Effect of Insulin Initiation on ED Return: A Mixed-Methods Study

Introduction and Objective: Medication changes during hospitalization create gaps in diabetes care, yet their interaction with social vulnerability is poorly understood. Predictive modeling identified insulin variables as top predictors of ED return, motivating this study. We examined whether social vulnerability moderates the effect of insulin initiation on ED return. Methods: We conducted a mixed-methods study at an academic medical center. Among 11,811 hospitalizations (6,460 patients with diabetes, 2022-2024), logistic regression examined three-month ED return with Social Vulnerability Index (SVI>85th percentile among the study population) as modifier. Ten provider interviews across four sites underwent thematic analysis. Results: Insulin addition reduced ED return (OR=0.75, p<0.01), but this effect was absent among high-SVI patients (OR=1.30, NS; interaction p=0.07) (Figure 1). Short-acting insulin at admission increased risk (OR=1.39, p<0.01). Provider interviews identified ten themes; cost barriers, health literacy gaps, and insulin-specific fears were most prevalent. Competing social priorities and documentation discrepancies further undermined transitions. Conclusion: Insulin initiation benefits low- but not high-SVI patients. Qualitative findings reveal system gaps explaining this disparity, supporting community-engaged interventions for socially vulnerable patients with diabetes. Disclosure S. Lee: None. F. Ovalle: Research Support; Current; Chiesi USA, Inc. Research Support; Ended; Lilly. Research Support; Current; JDRF, National Institutes of Health. A. Cherrington: Research Support; Current; Novo Nordisk.