Abstract
Abstract: :
Purpose: To evaluate accuracy of the Grand Seiko WR-5100K autorefractor, as compared to cycloplegic refraction with refinement in children. Methods: Subjects were 92 fifth- and sixth-grade children from a population with a high prevalence of astigmatism. Non-cycloplegic autorefraction measurements (based on 10 readings) were obtained with the WR-5100K while subjects viewed an ETDRS letter chart located at 4.0 m. Following non-cycloplegic autorefraction, subjects were cyclopleged (2 drops of cyclopentolate 1%) and refracted using the Nikon Retinomax K+ autorefractor. The cycloplegic autorefraction was then subjectively refined to determine the subject's best estimate of refractive error (BER). Analyses of accuracy compared measurements of the WR-5100K to each child's BER for spherical equivalent (M) and two Jackson crossed cylinder vectors (J0 and J45). Results: Results of cycloplegic refraction with refinement indicated that sphere values ranged from -6.00 D to +3.50 D (mean=-0.44, SD=1.85) and cylinder values ranged from 0 to +5.25 D (mean=+1.45, SD=1.58). The mean difference between measures with the WR-5100K and subject's BER was 0.18 D (SD=0.62), 0.004 D (SD=0.22), and -0.02 D (SD=0.15), for M, J0, and J45, respectively. The difference between the two measurements was significant for M, with WR-5100K measurements slightly less hyperopic than BER (t(91)=2.82, p=0.006). Differences between measurements for J0 and for J45 were not significant. Conclusion: The WR-5100K provided slightly less hyperopic measurements of spherical equivalent than cycloplegic autorefraction with refinement. Measurements of cylinder were in good agreement, even in this highly astigmatic sample of children.
Keywords: refraction • astigmatism