Purpose
Diabetic retinopathy (DR) is a leading cause of low vision and blindness; current DR classification was set by the Early Treatment Diabetic Retinopathy Study (ETDRS) using 7 stereoscopic retinal photographic fields taken by large tabletop cameras. While studies have investigated DR screening protocols using < 7 fields of view using tabletop cameras, none have investigated the ability of a handheld fundus camera to screen for DR. The purpose of this prospective study was to evaluate the feasibility of using a new FDA-approved, handheld, non-contact digital retinal camera, Pictor (Volk Optical Inc., Mentor, OH), to obtain and evaluate retinal images for the presence of DR compared to clinical exam findings.
Methods
Single fundus images of 56 adult diabetics (n=111 eyes) were acquired by a single trained imager pre- and post-dilation with the Pictor. Grading was performed by 5 masked ophthalmologists (1 general provider, 3 medical retina fellows, 1 medical retina attending) using a modified ETDRS grading system. Grading accuracy was compared to the dilated clinical examination findings obtained the same day of imaging. Of the images felt to be interpretable, the sensitivity and specificity of identifying low-risk (defined as no, mild, and moderate non-proliferative diabetic retinopathy (NPDR) and high-risk eyes (defined as severe NPDR and proliferative DR) for each grader was calculated compared to the clinical exam findings.
Results
Graders felt images were interpretable in 86-94% of pre-dilation photos and in 94-97% of post-dilation photos. Among the post-dilation photos, the sensitivity for identifying low- vs high-risk eyes ranged from 50-56% and specificity ranged from 92-98% compared to the clinical exam. Modifying the screening criteria by grading for the presence of moderate NPDR, severe NPDR, or PDR, the sensitivity for identifying high-risk eyes was 62-85% and the specificity was 72-90%.
Conclusions
The Pictor camera can take high quality fundus images that graders felt were of sufficient quality to screen for DR. While graders evaluating single fundus images acquired by Pictor are highly specific but not very sensitive in evaluating for high-risk DR, a modified screening criterion increased the sensitivity of identifying high-risk eyes with acceptable specificity. The Pictor camera may be useful in screening for DR and other pathology localized to the posterior pole.