Neighborhood Socioeconomic Deprivation and Adherence to Diabetic Retinopathy Screening Guidelines.

PURPOSE: Screening for diabetic retinopathy is imperative to prevent irreversible vision loss. Features to identify patient populations at higher risk for nonadherence to screening guidelines is key to improving visual outcomes. The purpose of this study was to evaluate the association between Area Deprivation Index(ADI) and adherence to diabetic retinopathy screening in a nationwide sample of patients. DESIGN: Retrospective cohort study METHODS: A twenty percent sample from fee-for-service Medicare claims from 1/1/16 to 12/31/22 was used. Medicare fee-for-service beneficiaries were eligible for inclusion in the study if they were enrolled in the plan as above and were determined to have a diagnosis of diabetes mellitus. Baseline demographic characteristics, comorbidities, and local healthcare resources were also ascertained. The primary exposure evaluated was ADI, a composite score incorporating factors related to the income, education, employment, and housing of a given neighborhood. The primary outcome evaluated was adherence to annual diabetic retinopathy screening, determined based on CPT codes for ophthalmic examination. A univariate and multivariate logistic regression model evaluating association between ADI and screening adherence was performed. A p-value less than 0.05 was considered statistically significant. RESULTS: Over 1 million Medicare beneficiaries were included in this analysis. Overall screening rates ranged between 51.7%(2022) to 55.3%(2016). When evaluating the screening rate by ADI subgroup, there was a consistent trend of decreasing screening rates with increasing ADI. Both univariate and multivariate regression models demonstrated a significant association between ADI and adherence to annual screening for diabetic retinopathy. The probability of having an annual screening were 24 percent lower in the subgroup of patients with ADI 86-100 compared to patients with ADI 1-15 (RR: 0.76, 95% CI: 0.76-0.77, p<0.001). A multivariate logistic regression model controlling for demographics, comorbidities, the Covid-19 pandemic, and county resources, yielded similar results. The probability of having an annual ophthalmic screening exam were 16 percent lower in the subgroup of patients with ADI 86-100 compared to patients with ADI 1-15 (RR: 0.84, 95% CI: 0.84-0.85, p<0.001). CONCLUSIONS: In this nationwide cohort, there was a significantly increased risk of nonadherence to annual diabetic retinopathy screening with increasing ADI.

Duke Scholars

Author Bradley Gordon Hammill Population Health Sciences

Author Durga Borkar Ophthalmology, Vitreoretinal Diseases & Surgery

Author Irene Song