Treatment of unilateral visual impairment on preschool vision screening

Editor—In their randomised controlled trial Clarke et al showed directly for the first time that treatment improves visual acuity, at least for moderate amblyopia.¹

However, we take issue with the conclusion that vision of 6/12 in one eye and 6/9 in the other should not be treated. There are several possible reasons why no effect of treatment was found for this group.

6/9 is likely to be within the normal range of results for this age group for the crowded test used rather than for a single letter test, in which 6/9 would be more likely to be subnormal at this age.

Some of these children may well have had simple uncorrected refractive error rather than amblyopia because there was no refraction before randomisation and treatment.

Test-retest data for children show that visual acuity varies by at least one line between different occasions.²,³

A ceiling effect at the mild end of the scale means that the effect size can be expected to be smaller in this group (less room for improvement). Given that the mild group probably contained several children with normal sight owing to misclassification and that the effect size would be expected a priori to be smaller in the mild group, the study probably did not have sufficient power to show an effect of treatment in this group. The fact that a significant interaction term was found between mild and moderate groups would be expected with the presence of a ceiling restriction on the effect size in the milder group. Furthermore, no data are provided on binocular fusion or stereopsis, outcomes which might have improved in the absence of change in visual acuity.

This is a worthwhile study on which the research community can build, but it cannot be expected to answer all the relevant questions and it would be premature to draw sweeping conclusions based on its findings.