Do picture-based charts overestimate visual acuity?

Comparison of kay pictures, lea symbols, HOTV and Keeler logMAR charts with Sloan letters in adults and children

Nicola S. Anstice, Robert J. Jacobs, Samantha K. Simkin, Melissa Thomson, Benjamin Thompson, Andrew V. Collins

Research output: Contribution to journalArticle

5 Citations (Scopus)
4 Downloads (Pure)

Abstract

Purpose Children may be tested with a variety of visual acuity (VA) charts during their ophthalmic care and differences between charts can complicate the interpretation of VA measurements. This study compared VA measurements across four pediatric charts with Sloan letters and identified chart design features that contributed to inter-chart differences in VA. Methods VA was determined for right eyes of 25 adults and 17 children (4±9 years of age) using Crowded Kay Pictures, Crowded linear Lea Symbols, Crowded Keeler logMAR, Crowded HOTV and Early Treatment of Diabetic Retinopathy Study (ETDRS) charts in focused and defocused (+1.00 DS optical blur) conditions. In a separate group of 25 adults, we compared the VA from individual Kay Picture optotypes with uncrowded Landolt C VA measurements. Results Crowded Kay Pictures generated significantly better VA measurements than all other charts in both adults and children (p < 0.001; 0.15 to 0.30 logMAR). No significant differences were found between other charts in adult participants; children achieved significantly poorer VA measurements on the ETDRS chart compared with pediatric acuity tests. All Kay Pictures optotypes produced better VA (p < 0.001), varying from -0.38 ± 0.13 logMAR (apple) to -0.57 ± 0.10 logMAR (duck), than the reference Landolt C task (mean VA -0.19 ± 0.08 logMAR). Conclusion Kay Pictures over-estimated VA in all participants. Variability between Kay Pictures optotypes suggests that shape cues aid in optotype determination. Other pediatric charts offer more comparable VA measures and should be used for children likely to progress to letter charts.

Original languageEnglish
Article numbere0170839
Pages (from-to)1-17
Number of pages17
JournalPLoS One
Volume12
Issue number2
DOIs
Publication statusPublished - 2017
Externally publishedYes

    Fingerprint

Cite this