Comparative study of the polaroid and digital non-mydriatic cameras in the detection of referrable diabetic retinopathy in Australia

Abstract

To show that the non-mydriatic retinal camera (NMRC) using polaroid film is as effective as the NMRC using digital imaging in detecting referrable retinopathy. A series of patients with diabetes attending the eye out-patients department at the Royal Victorian Eye and Ear Hospital had single-field non-mydriatic fundus photographs taken using first a digital and then a polaroid camera. Dilated 30 degrees seven-field stereo fundus photographs were then taken of each eye as the gold standard. The photographs were graded in a masked fashion. Retinopathy levels were defined using the simplified Wisconsin Grading system. We used the kappa statistics for inter-reader and intrareader agreement and the generalized linear model to derive the odds ratio. There were 196 participants giving 325 undilated retinal photographs. Of these participants 111 (57%) were males. The mean age of the patients was 68.8 years. There were 298 eyes with all three sets of photographs from 154 patients. The digital NMRC had a sensitivity of 86.2%[95% confidence interval (CI) 65.8, 95.3], whilst the polaroid NMRC had a sensitivity of 84.1% (95% CI 65.5, 93.7). The specificities of the two cameras were identical at 71.2% (95% CI 58.8, 81.1). There was no difference in the ability of the polaroid and digital camera to detect referrable retinopathy (odds ratio 1.06, 95% CI 0.80, 1.40, P = 0.68). This study suggests that non-mydriatic retinal photography using polaroid film is as effective as digital imaging in the detection of referrable retinopathy in countries such as the USA and Australia or others that use the same criterion for referral.