Kaleidos-based Binocular Vision Analyzer for the Estimation of Refractive Errors and Horizontal Ocular Deviation in Children

Clinical relevance: Knowing the agreement and interobserver variability of the 2WIN-S photorefractometer with cycloplegic retinoscopy and prism cover test is important for judging the clinical usefulness of the device.Background: This study aims to determine an agreement and interobserver reliability between 2WIN-S photorefractometer and cycloplegic retinoscopy for estimating refractive errors and between 2WIN-S and prism cover test for measuring horizontal ocular deviation in children.Methods: Two groups of 129 and 64 participants (range, 5 to 16 years) were recruited for the agreement and interobserver reliability studies, respectively. Non-cycloplegic 2WIN-S refraction was compared with cycloplegic retinoscopy for low and moderate myopia and hyperopia. Similarly, corneal reflex-function (CR-function) of 2WIN-S was compared with prism cover test for horizontal ocular deviation. Bland-Altman plots were used to depict agreement, and the intraclass correlation coefficient (ICC) was used for determining interobserver reliability of repeated measures.Results: The mean differences in spherical equivalent, J0 (Cartesian astigmatism) and J45 (oblique Jackson cross-cylinder) between cycloplegic retinoscopy and 2WIN-S were 1.10 (95% CI = 0.97 to 1.21, p < 0.001), -0.06 (95% CI = -0.09 to -0.03, p < 0.001) and 0.009 (95% CI = -0.01 to 0.03, p = 0.35), respectively. The overall median difference in the horizontal ocular deviation between prism cover test and 2WIN-S was -4.5 (Interquartile range = -12.0 to 4.0). For constant strabismus, this difference was insignificant (median= -2.0, interquartile range = -11.0 to 6.7, p = 0.19) between prism cover test and 2WIN-S. The ICC was high for the spherical equivalent (0.997; p < 0.001), J0 (0.932; p < 0.001) and J45 (0.901; p < 0.001), whereas it was low-to-moderate for horizontal ocular deviation (0.50; p < 0.001).Conclusion: 2WIN-S refraction was in agreement with cycloplegic retinoscopy for low and moderate myopia and hyperopia, and requires an adjustment for the consistent differences between cycloplegic retinoscopy and 2WIN-S refraction. The measurement of constant strabismus with 2WIN-S is comparable to that of prism cover test.