SP47. The Area Deprivation Index as a Predictor of Follow-up Adherence in Facial Fracture Patients: Addressing Socioeconomic Disparities in an Urban Emergency Setting

PURPOSE: Timely follow-up care is essential for patients with facial fractures to prevent long-term functional and aesthetic complications. However, adherence to follow-up remains poor, especially among socioeconomically disadvantaged populations. While the impact of individual socioeconomic factors on healthcare access has been extensively studied, the role of the Area Deprivation Index (ADI), a composite measure of neighborhood-level socioeconomic disadvantage, in predicting follow-up adherence for facial fracture patients has not been previously explored. This study addresses a significant gap by evaluating how ADI influences follow-up rates in an urban emergency department (ED) and examining interactions between ADI and demographic factors, such as race, age, and insurance status, using advanced statistical and machine learning models. METHODS: A retrospective cohort study of 629 facial fracture patients treated in an urban ED was conducted. We collected demographic data, socioeconomic indicators (including ADI and insurance status), and clinical outcomes. Propensity score matching (PSM) was utilized to balance ADI groups based on covariates. Subgroup analyses were performed to examine interactions between ADI and key demographic variables. Multilevel logistic regression was employed to account for patient clustering by ZIP code, and machine learning models (Random Forests, Support Vector Machines) were applied to identify key predictors of follow-up adherence. Model performance was assessed using accuracy and area under the receiver operating characteristic curve (AUC). RESULTS: The mean age of the cohort was 36.16 years (SD = 21.24), with 72.34% male and 50.08% identifying as Black or African American. Overall, 51.93% of patients completed follow-up. Higher ADI scores were significantly associated with lower follow-up completion rates (OR = 0.997, 95% CI: 0.995-0.999, p < 0.001). Subgroup analysis revealed that pediatric patients (OR = 0.993, 95% CI: 0.986-0.999, p = 0.04) and those without insurance (OR = 0.983, 95% CI: 0.972-0.994, p = 0.001) were particularly vulnerable to the effects of socioeconomic deprivation on follow-up adherence. Multilevel modeling confirmed the negative association between ADI and follow-up, with age positively associated with adherence (OR = 0.997, 95% CI: 0.995-0.999, p = 0.01). Machine learning models supported these findings, with the Random Forest model achieving an AUC of 0.644 and the Support Vector Machine model an AUC of 0.615, identifying ADI and insurance status as the strongest predictors of follow-up non-adherence. CONCLUSIONS: This is the first study to demonstrate that neighborhood-level socioeconomic disadvantage, as measured by ADI, is a significant predictor of follow-up non-adherence among facial fracture patients in an urban ED. Higher ADI scores were associated with lower follow-up rates, particularly in pediatric and uninsured populations. These results emphasize the need for targeted interventions to address socioeconomic disparities in post-ED care. The use of machine learning models further underscores the potential of predictive analytics in identifying at-risk populations and informing healthcare interventions aimed at improving continuity of care in disadvantaged communities. Future research should develop strategies to mitigate these barriers and enhance access to follow-up care.