Validity of non‐mydriatic cameras for screening and follow‐up in diabetic retinopathy

Abstract

Abstract Purpose: To determine the validity of a non‐mydriatic camera for screening and grading diabetic retinopathy (DR). To establish the number of photographs and the field width needed for a correct DR follow‐up. Methods: A cross‐sectional, observational study was carried out to assess the validity of the non‐mydriatic Topcon TRC‐NW6S retinograph. Validity proportions were calculated. Kappa analysis was made to determine the agreement with conventional fundoscopy exploration performed by indirect ophthalmoscopy and retinal biomicroscopy. One 45º single‐field non‐mydriatic digital photograph was taken in 82 eyes for DR screening. For DR grading, several combinations of retinal fields were photographed in 247 eyes, first without pupillary dilatation and later with mydriasis. Results: In DR screening, 88.2% sensitivity and 96.9% specificity were obtained, where 9% of the tests were invalid. In DR grading diagnosis, the kappa analysis showed close agreement (k>0.8) based on at least two 45º photographs with mydriasis. However, when attempting to detect macular edema (ME), the maximum kappa statistic obtained did not go above 0.71, showing 67% maximum sensitivity. The sensitivity for detecting derivable DR was similar to that obtained with indirect ophthalmoscopy (94‐98%). Conclusions: The non‐mydriatic retinograph is a valid instrument for DR screening only when taking one 45º non‐mydriatic photograph per eye. However, given that the sensitivity for proliferative DR (PDR) was worse, when grading DR, we would recommend obtaining nine retinal photographs (mosaic) with mydriasis. Used in this way, the apparatus is extremely useful for detecting derivable DR cases.