In 1981, when the Kay Picture Test was conceived, the picture tests used in the assessment of children's visual acuity were universally regarded as inaccurate and considered only as a guide or comparison of sight in the two eyes. Hazel Kay sought to produce pictures that accurately measured visual acuity in the same way as Snellen's letters and that young children could easily recognise.
Over a two-year period more than 60 pictures were drawn according to the physiological principles employed in Snellen's letter construction: the component parts of a construction grid subtend 1 minute of arc at the nodal point when viewed at the correct distance. The more complex structure of pictures required a 10 x 10 construction grid rather than the 5 x 5 grid used with letters.
156 children between the ages of 19 and 42 months were asked to identify each picture. 24 of these pictures were instantly recognisable to children of this age. Two other pictures have since been eliminated as having lost their appeal.
In a comparative study (Kay, 1984) between the linear Snellen test and each single picture (N=160) the correlation coefficient was 0.9043. The Kay Picture Test contains only those pictures proven to be both recognisable and accurate.
The Screening Book was designed after it became clear that children were able to match with pictures at a much earlier age than with unfamiliar shapes or letters (approx 6-12 months earlier) and a matching card helped shy, under-confident children. The eight pictures were chosen for three reasons:
- They rated near the top in the accuracy and recognition research results.
- The pictures have counterparts with similar outlines which may be confused at the limit of visual acuity (i.e. as in C and O, and H and N in letter testing).
- Eight pictures provides a sufficient variety to maintain interest while still being a small enough number to use with matching. Compared to tests using only four symbols, the probability of a random guess being correct is havled.
The Crowded LogMAR test is designed to enable linear acuity measurement before letter tests are possible. By introducing the crowding effect it allows for the earlier detection and treatment of amblyopia.
Single optotype tests do not always identify small degrees of amblyopia and a linear acuity test is advisable as soon as possible. Illiterate pre-school children perform better and more reliably with naming familiar pictures than with matching letters or symbols. Matching unfamiliar objects requires preparation before the test to practice and confirm that the child understand the procedure.
By three years of age most children have the necessary concentration to achieve accurate reliable results with the Crowded LogMAR Kay Picture Test. Familiar pictures are regarded as easy by children aged three or more and the test will be quicker and more fun for the child. Excellent correlation with LogMAR letter tests have been established (see bibliography).
The following values are based on a retrospective study of children discharged from West Hertfordshire PCT Orthoptic clinic with normal vision. Children included in this study were all established as having normal bifoveal single vision and normal uniocular acuities of 6/6 (0.00 LogMAR) or better with a letter test and/or confirmed with a normal reading on refraction.
Uniocular visual acuity measurements using the Kay Picture Test Crowded pictures were recorded at first testing, together with the child's exact age and gender. The visions of 224 children aged 2 years 2 months to 5 years 10 months were retrieved. There were 98 girls and 126 boys.
The results were divided into two groups, under four years of age and over four years. The age groups were chosen to reflect the likely difference in concentration levels under and over four years which will affect recorded acuity measurements and to tie in with the government proposed screening age.
The range of acuities was similar in both groups:
|under 4s||-0.100 to 0.200 LogMAR (6/5 to 6/9.5 or 20/15 to 20/30 Snellen)|
|over 4s||-0.100 to 0.100 LogMAR (6/5 to 6/7.5 or 20/15 to 20/25 Snellen)|
Under four Years (n=118)
- The mean age was 3y 6m. Only 10 children were under 3 years of age.
- 111 children (94%) had visions of 0.100 LogMAR (6/7.5 or 20/25 Snellen) or better in both eyes.
- 96 children (81%) achieved exactly the same vision in each eye.
- 112 children (95%) had a difference of no more than 0.050 (two pictures) between right and left acuities.
Four and Five Year olds (n= 106)
- 97 children (92%) had visions of 0.050 LogMAR (6/6 or 20/20 part Snellen) or better in both eyes.
- 97 children (92%) achieved exactly the same vision in each eye.
- 101 children (95%) had a difference of no more than 0.025 (one picture) between right and left acuities.
Normative values for children under 4 years of age are:
- Uniocular acuities of 0.100 LogMAR (6/7.5 or 20/25 Snellen) or better with an intraocular difference of no more than 0.050 LogMAR (two pictures).
Normative values for children aged 4 and five years are:
- Uniocular acuities of 0.050 LogMAR ( 6/6 or 20/20 part Snellen) or better, with an intraocular difference of no more than 0.025 LogMAR (one picture).
- Kay H, 1983, New method of assessing visual acuity with pictures, British Journal of Opthamology, 67, pp 131-133.
- Kay H, 1983,The Optician, 13 May.
- Kay H, 1984, A new picture visual acuity test, British Orthoptic Journal, 41, pp 77-80.
- Jones D, Westall C, Averbeck K, Addolell M, 2003, Visual acuity assessment: A comparison of two tests for measuring children's vision, Ophthalmic and Physiological Optics, 23(6), pp. 541-546. Abstract available online.
- Elliott M C, Firth A, 2007, The LogMAR Kay Picture Test and the LogMAR Acuity Test: A comparative study, Eye, 27 Sept. Full-text available online.
- Cross M, Ager V and Fakis A, 2010, The range of visual acuity using the crowded Kay Picture test and the range of refractive error present in children aged 42.48 months. P66 The British and Irish Orthoptic Journal 2010 vol 7.