Comparison of spherical equivalent refraction and astigmatism measured with three different models of autorefractors.

PURPOSE: The purpose of this study was to compare refractions measured with three different autorefractors.
METHODS: The refractive error of each eye of 50 adults aged 17 to 59 years (mean, 30.5 years) was measured without cycloplegia using the Canon R-1 and two newer instruments, the Grand Seiko WR-5100K and the Nidek ARK 700-A. For the first two, an isolated line of 20/100 letters on an ETDRS chart at 4.0 m served as a target, whereas for the Nidek, the subject looked at a picture of a balloon in the instrument. Five readings were taken for each eye, and the data (sphere, negative cylinder power, and axis) were analyzed using Fourier decomposition of the power profile. Each reading was broken down into the spherical equivalent (M) and two Jackson crossed-cylinder vectors, J0 and J45. Right-eye results are reported.
RESULTS: The mean spherical equivalent refraction measured by the Canon R-1 was -2.44 D. Measurements from the Grand Seiko were more hyperopic (mean M, -2.01 D), whereas those from the Nidek were more myopic (mean M, -2.66 D). Correlation of M for each pair of autorefractors was 0.99. For J0, the Canon was more minus than the other two instruments by 0.15 D compared with the Nidek and 0.13 D with the Grand Seiko, and on this component, the correlation of the Canon with each of the other two was 0.87. Mean J0 values for the Nidek and Grand Seiko were similar, 0.05 D and 0.03 D, respectively, and mean J45 values were 0.04 D for both instruments. Correlations were 0.97 between these two autorefractors for each of the two components. J45 measured by the Canon was more positive than the other two by 0.06 D. For J45, the correlation of the Canon with the Grand Seiko was 0.40 and with the Nidek was 0.38. In 92% of the eyes, the absolute difference in cylinder power between the Grand Seiko and the Nidek was < or =0.25 D. Only 42% of the differences between the Canon and the Grand Seiko and 40% of the differences between the Canon and the Nidek were this small.
CONCLUSIONS: The Canon provided more myopic readings than the Grand Seiko and more hyperopic readings than the Nidek. The Canon measured more astigmatism that did not correlate well with the other instruments, whereas measurements of astigmatism taken by the Nidek and the Grand Seiko showed good agreement. If the Grand Seiko is used in place of the older Canon, the differences in spherical equivalent and astigmatism must be considered.