Evaluation of the Accuracy of Estimation Retinoscopy

Background: Some children are unable to cooperate for retinoscopy because they object strongly to the placement of lenses close to their faces. For these children, it would be ideal to obtain an accurate estimate of refractive error without using lenses. Techniques of estimation retinoscopy include sliding the sleeve of the Copeland retinoscope downward or moving closer to the patient until neutrality is achieved. The purpose of this study was to evaluate the accuracy of estimation techniques by comparing results to standard retinoscopy with loose lenses in cooperative children. Methods: A Copeland retinoscope was calibrated using a schematic eye and loose lenses. A scale was created adjacent to the sleeve of the retinoscope which allowed an estimate of refractive error based on the position of the top of the sleeve. Estimation retinoscopy followed by standard retinoscopy with loose lenses was done on 100 eyes of 50 children after cycloplegia. Results: Estimation of spherical equivalent for myopia less than 4 D and hyperopia less than 2 D correlated strongly with results obtained by standard retinoscopy with loose lenses (r = 0.87). Estimation retinoscopy had a sensitivity of 88%, specificity of 67%, positive-predictive value of 58%, and negative-predictive value of 92% in the detection of amblyogenic refractive errors. Conclusions: Estimation retinoscopy has very good accuracy for low levels of myopia, hyperopia, and astigmatism. Techniques of estimation may be useful in excluding amblyogenic refractive errors, particularly in children who object to loose lenses held close to them.