Abstract
Purpose :
Knowledge about the shape of the psychometric function for visual acuity charts allows for better modelling of vision testing and more efficient estimation of acuity thresholds. We assessed whether 4 commonly used monitor based chart systems gave different shaped psychometric functions for child participants.
Methods :
Data were obtained from 22 children, aged between 38 and 71 months (mean 54), with acuity better than 0.32 logMAR on LEA or EVA testing. For each child acuity measurements were attempted twice on right and left eyes for each of 4 commercially produced pictogram charts, presented on LED monitors: VistaVision Visual Acuity Tester; Thomson Test Chart; Rodenstock Phoromat 2000; and Optos Smart Chart (optotype sets for each, Figure 1). Measurements were analyzed if the maximum correct on one line was 80% or more and the minimum correct on one line was 40% or less. Probit analysis was performed on each measurement’s data set (logMAR units), with the upper asymptote set for a 1 % lapse rate and the lower asymptote based on the number of alternatives for each optotype set (Vista 25%, Thomson 12.5%,Rodenstock 16.7%,Optos 11.1%). For each measurement a threshold acuity was obtained and a Probit size (an indication of the flatness of the psychometric function). For each participant the acuity and Probit size were averaged across eyes and sessions (i.e.based on up to four measurements).
Results :
Different charts yielded different shaped psychometric functions. Probit sizes differed for different charts, (F3,39=8.198,p<0.001) averages being (from steepest to flattest ): Thomson 0.05logMAR, SD0.024; Optos 0.10logMAR, SD0.04; Vista 0.13logMAR, SD0.05; Rodenstock 0.15logMAR, SD0.11. Acuity values also differed between chart types (F3,39=51,p<0.001): averages Thomson-0.04 logMAR, SD0.12; Optos 0.08logMAR, SD0.12; Vista 0.13logMAR, SD0.08; Rodenstock 0.24logMAR, SD0.13.
Conclusions :
The chart design with the smallest Probit values (Thomson) gives a similar shaped psychometric function to that obtained previously with letter charts in adults. Other charts give much flatter functions, indicating unreliable and inefficient acuity measurements. Optotype selection matters for pediatric acuity measurement.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.