2251-P: Divergent Associations of Area Deprivation Index with Glycemic Control and Diabetic Eye Screening in Veterans

Introduction and Objective: Socioeconomic factors measured by area deprivation index (ADI) are associated with higher rates of diabetes and lower rates of optimal diabetes care in non-VA settings. We evaluated association of ADI with A1c and retinal screening within the VA healthcare system. Methods: Veterans with type 2 diabetes without diabetic retinopathy who attended a primary care appointment within the VA between 4/1/2023 and 3/31/2024 were included. Veteran address during the study year was mapped to ADI based on census block (higher ADI indicates greater socioeconomic deprivation). Having well-controlled diabetes (HbA1c <8% most recent in 18 months prior to 3/31/2024) and retinopathy screening were co-primary outcomes. Veterans completing an eye exam through a VA eye clinic, teleretina screening program, or VA community care within 18 months after PCP visit were considered screened. Logistic regression modeling with sandwich estimators to account for medical center clustering was used to estimate associations of ADI with outcomes, adjusting for age, gender, race, ethnicity, and number of PCP visits (only for screening outcome). Results: The cohort (n=1,990,314) was primarily male (92%), White (68.2%), non-Hispanic or Latino (89%), and mean age of 68 years. Overall eye screening rate was 77%. With increasing ADI quintile, the adjusted odds ratio (OR) for A1c <8% decreased (OR 0.94, 0.90, 0.87, and 0.83 for 2nd, 3rd, 4th, and 5th quintiles vs 1st quintile, all p<0.001). Conversely, increasing ADI quintiles were associated with greater odds of diabetic eye screening (OR 1.11, 1.22, 1.36, and 1.44 respectively, all p<0.001). Conclusion: Within the VA, greater socioeconomic deprivation was associated with worse glycemic control but higher rates of eye screening. Exploring factors associated with retinal screening, whether range of screening modalities, differing non-VA insurance options, or recognition of at-risk populations, may inform strategies for improved comprehensive diabetes care. Disclosure K. Christopher: None. G. Deitz: None. S. Raghavan: None. J. Reusch: Advisory Panel; Current; Medtronic. Funding VA-ORD (1IK2CX002805-01A1)