Agreement Between Non-Cycloplegic Photorefraction and Retinoscopy in Pediatric Refraction

Abstract

Accurate assessment of refractive error in children is essential for clinical decision-making, yet agreement between non-cycloplegic techniques remains uncertain, particularly due to differences in accommodative demand. This study evaluated the agreement between static retinoscopy and handheld photorefraction for measuring spherical power, cylindrical power, and spherical equivalent in children aged 4–16 years and assessed whether agreement varied by refractive status. In this cross-sectional observational study, 193 children underwent objective refraction during a single visit using non-cycloplegic static retinoscopy (distance fixation) and handheld infrared photorefraction (~1 m fixation). Inter-method differences were analyzed using Bland–Altman plots, intraclass correlation coefficients (ICCs), mean absolute error (MAE), and non-parametric tests. Photorefraction showed a statistically significant myopic shift compared with retinoscopy for spherical power (−0.16 D), cylindrical power (−0.24 D), and spherical equivalent (−0.28 D). Agreement was moderate in statistical terms for spherical equivalent (ICC = 0.73) and spherical power (0.71), and lower for cylindrical power (0.46); however, wide limits of agreement indicate clinically relevant variability. MAE for spherical equivalent was 0.80 D overall, with 45.1% of measurements within ±0.50 D, and varied by refractive status, being lowest in emmetropic eyes and higher in hyperopic and myopic eyes. These findings indicate that, under non-cycloplegic conditions, photorefraction shows modest mean differences but substantial individual variability, likely influenced by differences in accommodative demand between techniques. While suitable for pediatric vision screening, photorefraction should not be considered interchangeable with retinoscopy for individual refractive assessment.