The Path to Improving Diabetic Retinopathy Outcomes: Evaluating Neighborhood Characteristics and Healthcare Quality

Duke University, Durham NC

Investigators

Abstract

Abstract Diabetic retinopathy is a leading cause of preventable, irreversible blindness in the United States. The risk of blindness from diabetic retinopathy can be minimized, but relies on a) appropriate screening for early identification of DR, b) regular monitoring once DR has been identified, and c) appropriate treatment as DR progresses to more advanced stages. At least one third of patients with diabetes in the United States do not receive annual dilated eye exams for screening, as recommended by the American Academy of Ophthalmology guidelines. Further, up to 25% of patients with identified DR do not receive appropriate monitoring after diagnosis. The central hypothesis of the proposed work is that neighborhood and community characteristics, encompassed in the area deprivation index (ADI), will be as or more important than individual patient characteristics for explaining nonadherence to DR screening and monitoring. The specific aims of our study are to 1) Evaluate the impact of patient, neighborhood, and community characteristics on patterns of screening for DR among Medicare beneficiaries with diabetes mellitus and to 2) Evaluate the impact of patient, neighborhood, and community characteristics on patterns of follow-up and incidence of vision-threatening complications from DR. Using Medicare data linked with ADI and measures of local healthcare environment characteristics, we will use a repeated measures modeling strategy to (1) identify factors associated with adherence to recommended screening intervals over time within a population of patients at risk for developing DR and to (2) identify factors associated with adherence to recommended monitoring intervals over time with a population of patients with DR. We will also use a time-to-event modeling strategy to identify factors associated with the development of DR complications. Results based on our proposed work with nationwide health data could be used to inform quality measures and reimbursement incentives, decrease cost of care, and improve health equity by better allocating resources to decrease the rising incidence of preventable, irreversible blindness in the 10 million people in the US with diabetic retinopathy.