Abstract
Purpose :
Computerised smart charts are used in practice to measure visual acuity in pre-literate children; however, little is known about the accuracy and repeatability of many of the picture optotypes available. This study firstly compared the ability of four computerised smart charts to measure visual acuity in pre-literate children with two gold standard paediatric visual acuity tests (EVA and LEA). Secondly, the repeatability of different charts across visits was evaluated.
Methods :
Unaided right and left eye acuities were measured twice (separate visits) with four computerised picture charts (Optos, Thomson, Rodenstock, VistaVision), the EVA and a printed LEA logMAR chart in children aged 3 to 5 years. Chart and eye order were randomised for each child each visit. The computerised charts were presented at a six-metre test distance (mirror), the EVA and LEA charts were tested at 3 metres. Visual acuity testing was stopped once three errors were reported on a line; scoring was by letter-counting method and recorded in logMAR.
Results :
Twenty-four children were recruited (age: 54.3±9.8 months; 11 males). No child had distance spectacles and mean spherical equivalent refraction was 0.36±0.58D, range: -1.00-+1.50D. Two children (58 and 59 months of age) were uncooperative and unable to be tested on any chart. For the remaining 22 children, right and left acuity measures were highly correlated at the first visit (r-values ranging from 0.75-0.94, p<0.01); subsequently, right eye measures were used for remaining analyses. Mean EVA acuity at the first visit was -0.02±0.10 logMAR, range: -0.10-0.20; Fig 1 plots acuities for all charts (visit 1). Repeated-measures ANOVA showed significant main effects of chart (p<0.01) and visit (p=0.03). Post-hoc analysis (paired t-tests) demonstrated significant differences between all computerised charts and the EVA (p≤0.01); visit 2 elicited improved visual acuity results on the EVA, Thomson and Optos charts (p<0.05). Bland-Altman analysis demonstrated wider 95% limits of agreement for the computerised charts (compared to EVA) with the Thomson chart showing the least variability.
Conclusions :
All four computerised charts resulted in poorer visual acuity measures as compared to the gold standard, EVA. This has important implications on clinical practice; care should be taken when interpreting visual acuity measures from picture chart displays on computerised smart charts.
This is a 2020 ARVO Annual Meeting abstract.