Abstract
Purpose :
Telemedicine programs to assess diabetic retinopathy (DR) are a practical screening method to reach at-risk urban communities with poor access to specialized healthcare. We performed a retrospective chart review to evaluate the severity of DR and the rate of positive telemedical screens resulting in completed ophthalmology appointments among Temple University Hospital (TUH) patients.
Methods :
We evaluated a predominantly underserved African American patient cohort, 18 years and older, carrying a diagnosis of Diabetes Mellitus with no recent dilated fundus examination who underwent fundus photography screening for DR in TUH primary care clinics between March 2016 and May 2017. We investigated the percentage of the population that exhibited previously undetected DR, the relative severity of disease, and the follow-up rate with a specialist following a positive screen.
Results :
We conducted 689 retinal screening exams. The average HgbA1c among these patients was 8.05% (range, 4.2 – 18.1%). Three hundred forty-three (49.8%) of 689 patients were found to have no DR in either eye while 103 patients (14.9%) were found to have some level of DR in at least one eye. Of these patients, 33 (4.8%) had at least one image graded as “unspecified” DR and 70 (10.2%) patients had specified grades in both eyes: 48 (68.6%) had mild DR, 9 (12.9%) had moderate DR, 11 (15.7%) had severe DR, and 2 (2.9%) had proliferative DR in one or both eyes. Due to unassessable photos, we were unable to negatively or positively screen 243 (35.3%) patients for DR. Based on other findings in the retinal photographs, 67 (9.7%) patients were suspected to have another ophthalmic condition. Among the 689 screening exams, 357 (51.8%) resulted in a request for a referral to ophthalmology for any level of DR, inability to assess, or suspicion of another diagnosis. Of the 357 referral requests, 61 (17.1%) appointments with the ophthalmology clinic were scheduled. Thirty-four (55.7%) patients attended the appointment while the other 27 (44.3%) either no-showed or cancelled.
Conclusions :
Mere identification of referral-warranted DR is insufficient. In order to properly deliver necessary treatment to at-risk patients, a successful telemedicine screening program must close the communication gap between screening and diagnosis by reviewer to provide timely follow-up by eye care specialists.
This is a 2020 ARVO Annual Meeting abstract.