Abstract
Purpose: :
To determine the ability to detect optic nerve findings suggestive of glaucoma in patients undergoing telemedicine assessment for diabetic retinopathy in the primary care environment with a digital imaging system and remote reading center.
Methods: :
Readers trained and certified to assess images, obtained with the DigiScope in the primary care setting, for the presence of diabetic retinopathy at a telemedicine reading center were provided additional training on glaucoma and the optic nerve abnormalities associated with glaucoma by literature and a working session with a glaucoma expert. Readers were instructed to classify cup-to-disc ratios as 0 to 0.5, 0.6 to 0.8, and greater than 0.8 (based on levels of suspicion for glaucomatous optic nerve changes). Random images of the optic discs of patients being evaluated for diabetic retinopathy at a telemedicine reading center were assessed by a glaucoma expert. Agreement between readers and the glaucoma expert for cup-to-disc ratio and asymmetry was reviewed.
Results: :
Over a 3 month period, the images of both eyes of 30 random patients were reviewed with classification of cup-to-disc ratios. Agreement between the readers and glaucoma expert was excellent with complete agreement in 26 out of 30 patients. The cup-to-disc ratio was overestimated by the readers in 3 cases and underestimated in only one case. Asymmetric optic disc cupping was noted in 2 cases with agreement between the readers and glaucoma expert.
Conclusions: :
Images obtained with telemedicine systems designed for diabetic retinopathy assessment may reveal findings not associated with diabetic eye disease. With additional training, readers are able to detect optic nerve findings suggestive of glaucoma allowing for appropriate referral for further evaluation. Excellent agreement was noted for classification of optic nerve cup-to-disc ratio and asymmetry between trained readers and a glaucoma expert.
Keywords: clinical (human) or epidemiologic studies: systems/equipment/techniques • imaging/image analysis: clinical • diabetic retinopathy