ACCURACY AND PRECISION OF THE TOMEY VIVA INFRARED PHOTOREFRACTOR

Purpose. The Fortune Optical (Tomey VIVA) VRB-100 video refractor was tested to determine its accuracy and precision in measuring manifest r efractions of human eyes with and without cycloplegia. The specific is sues addressed included its accuracy in measuring spherical and cylind rical refractive errors and its precision in refracting near emmetropi a. Methods. To determine its ability to measure moderate to high (>4.0 0 D) myopia, we compared the ViVA's refractions to those taken by a Ca non Autorefractor R1 and a retinoscopist. A spherical lens series from -7.00 to +7.00 D at 1.00 D intervals, or -5.00 to +5.00 at 0.50 D int ervals, was placed over a subject's eye, which was then covered by an infrared (IR) filter, refracted, and analyzed to determine the ViVA's ability to measure spheres. Subjects with refractive errors of -2.00 t o +2.00 DS (diopters sphere) and 0 to 1.00 DC (diopters cylinder) were refracted 7 to 15 times during 1 sitting to determine the ViVA's prec ision. The instrument's accuracy in measuring cylinders was tested by placing +3.00 to -3.00 D cylinders (at 0.50 D intervals) over the eyes of subjects at 0 degrees, 45 degrees, 60 degrees, 75 degrees, 90 degr ees, 105 degrees, 120 degrees, and 135 degrees. Results. The ViVA meas ured spheres of +/-3.00 D with a root mean squared (rms) error of 0.5/-0.1 D. Beyond this power, its accuracy progressively worsened. In so me subjects, irregular intensity profiles compromised the ViVA's accur acy even with low spherical refractive errors. The VIVA was very preci se in measuring spheres from +2.00 to -2.00 D and cylinders from 0 to 1.00 D. Although the VIVA adequately measured ail cylinder powers at 0 degrees and 90 degrees, the accuracy of cylindrical power measurement decreased with obliquity; only cylinders less than or equal to 1.00 D magnitude were accurately measured at 45 degrees and 135 degrees. Con clusions. We conclude that, although the ViVA offers many attractive f eatures for vision screening, it is seriously limited by its inability to properly detect and measure oblique astigmatic errors.