The aim of this study was to compare the consistency of spherical diameter, column diameter by VX120 multi-diagnostic unit and subjective refraction.

The refractive surgery patients (71 cases of 142 eyes) from March 2018 to April 2018 in Beijing Tongren Hospital affiliated with Capital Medical University were collected, and analyzed differences between the results from VX120 multi-diagnostic scotopia mode and bright mode of objective optometry, and subjective optometry correction comprehensive refractometer, including spherical mirror photometry (S), column lenticular luminosity (C), equivalent spherical mirror luminosity (SE), column mirror luminosity with horizontal and vertical level (J0) and column mirror luminosity with 45°(J45). The refraction data were corrected by Rosner-Glynn-Lee method. The clustered signed rank test was used to compare the measurements of different devices. If there was not shown statistical difference, the consistency using intra-class correlation coefficient and the Bland-Altman 95% limits of agreement were performed.

S measured by VX120 with clear vision and VX120 with dark vision were -5.63 (-7.25, -4.25) and -5.75 (-7.00, -4.25), respectively. C measured by VX120 with clear vision and VX120 with dark vision was -0.75 (-1.50, -0.50) and -0.75 (-1.50, -0.50), respectively. SE measured by VX120 with clear vision and VX120 with dark vision was -6.19 (-8.09, -4.63) and -6.20 (-7.84, -4.53), respectively. J0 measured by VX120 with clear vision and VX120 with dark vision was -0.02 (-0.22, 0.21) and 0.03 (-0.15, 0.30), respectively. J45 measured by VX120 clear vision and VX120 dark vision was 0.04 (-0.22, 0.34), 0.04 (-0.22, 0.34), respectively. S, C, SE, J0 and J45 measured by subjective optometry were -5.50 (-7.00, -4.25), -0.50 (-1.00, -0.25), -6.02 (-7.50, -4.50), 0.00 (-0.21, 0.10), and 0.00 (-0.21, 0.10), respectively. S, C, J0 and J45 measured by CANON were -5.62 (-7.37, -4.50), -0.62 (-1.12, 0.00), 0.00 (-0.20, 0.03) and 0.00 (-0.10, 0.08), respectively. There was no statistic difference between VX120 with clear vision and subjective optometry for J0 and J45 (Z=-0.11, 1.64; P>0.05). The interclass correlation coefficients with 95% confidence intervals of J0 and J45 was 0.92 (95%CI=0.88, 0.96) and 0.93 (95%CI=0.89, 0.97), respectively. In addition, Bland-Altman plots showed that data points were in the confidence intervals, indicating that parameters measured by subjective optometry and VX120 were in accord with each other. There was no statistic difference between VX120 with dark vision and subjective optometry for S and J45 (Z=1.71, -0.24; P>0.05). The interclass correlation coefficients with 95% confidence intervals of S and J45 was 0.99 (95%CI=0.98, 1.00) and 0.90 (95%CI=0.86, 0.94) respectively. What′s more, Bland-Altman plots showed that data points were in the confidence intervals, indicating that parameters measured by subjective optometry and VX120 with dark vision were in accord with each other.

The difference between the optometry results from VX120 with dark vision and VX120 with bright vision mode and the subjective optometry was within the clinical acceptable range, which could provide an effective objective basis for the subjective optometry. There was no significant difference between VX120 and objective optometry, two devices could be considered clinically equivalent for refractive measurement.