Establishing Prediction Intervals for the SpeedWheel Acuity Test in Adults and Children.

Abstract

The SpeedWheel (SW) test is an objective test of visual acuity (VA) using suppression of the optokinetic nystagmus (OKN). Here, we established prediction intervals of the SW measures compared to Snellen acuity in adults and children. In this prospective, single center study, subjects aged at least 4 years underwent testing of VA with SW, Landolt-C, and Tumbling-E symbols (Freiburg acuity test: FrACT-C, FrACT-E). Prediction intervals were established for SW compared to FrACT-C or -E and for FrACT-E compared to FrACT-C. Mixed linear effect models were applied for statistical analysis. From 241 subjects, 471 eyes were included: median age 36 years, range 4 - 88 years, 43.6% male, 56.4% female. Eyes included were either healthy or had various underlying ophthalmic conditions. Prediction intervals for SW to estimate FrACT-C or -E acuity showed a similar range compared to the prediction interval of FrACT-C for the estimation of FrACT-E acuity. For each acuity step, there was no influence of age. Up to an SW acuity of 0.7 logMAR, 80% of the subjects had a FrACT-C acuity that was at most 1.6 logMAR lines below, and for an SW acuity of 1.0 logMAR, FrACT-C acuity was not worse than 4 logMAR lines. Prediction intervals for eyes with refractive error, cataract, visual field loss and retinal disease did not differ significantly from healthy eyes in contrast to eyes with amblyopia or multiple ophthalmic disorders. SW correlated well to FrACT tests and results of a previous study fell within our prediction estimates. Our prediction intervals for SW acuity may be used to estimate Snellen acuity (FrACT-C and -E) in the clinic in adults and children unable to cooperate in other acuity testing.