MON-707 Improving Screening for Diabetic Retinopathy in a Resident Based Clinic

Abstract Introduction Diabetic retinopathy is the leading cause of blindness in US Adults. In order to improve screening rates, we partnered with the Division of Ophthalmology and installed an onsite retinal camera at our primary care clinic. This led to an improvement in EMR reported screening rates from 20.5% to 44% over the first 3 months. We noticed that any fundus photo, whether gradable or not, led to an automatic annotation in EMR (EPIC) health Maintenance that screening had been completed. Abnormal or ungradable (quality too poor to interpret) retinal photos must be followed up with a complete ophthalmologic evaluation. We designed a chart audit to investigate further whether ungradable retinal photos were being followed up appropriately. Methods A retinal camera was installed in the clinic, and patients obtained DR screening during their routine visits from May through October 2018. The nursing staff received training on using the camera and ensuring image quality. These images were then sent to an Ophthalmologist and resulted within the work week. Patients with an abnormal or poor-quality retinal photo were contacted by their resident PCP. We did a retrospective chart review of patients with ungradable photos evaluating whether patients were contacted and whether they followed up with Ophthalmology in the 3 month period after the initial intervention. Results Of the 131 patients who received fundus photos in the study period, 29 (22%) had ungradable photos. Twenty-four of these patients were contacted and ophthalmology consults were placed for 22 patients. Eleven (38%) of these patients went on to complete screening with Ophthalmology within 3 months of the ungradable photo. Eighteen patients, or 62% of ungradeable photos, remained incorrectly identified as having completed retinopathy screening by EMR. Discussion Over reliance on EMR reporting features can lead to incorrect assumptions about DR screening. Based on this analysis, we need to design better interventions for following up on ungradable photos and ensuring appropriate follow up. One such intervention may be changing how EMR reports ungradable photos. EPIC is a widely used EMR in outpatient settings and other practices may be facing similar issues.