Skip navigation

Pre-school vision screening

Screening issues
Systematic reviews

Bandolier has been asked about the value of testing children for amblyopia, defined as a unilateral or, rarely, bilateral decrease of vision for which no obvious cause can be found on physical examination of the eye. The aim of pre-school vision testing in pre-school children is to detect a range of conditions, including amblyopia, refractive errors and non-cosmetic squints which cannot be detected without screening.

Screening issues

The basic principles of screening are that:

Are these conditions met for pre-school vision screening?

Systematic reviews

A systematic review with a heroic search strategy [1], say that they are not.

It found no studies that could document the natural history of squint, amblyopia or refractive errors in three to four year old children. What little information is available argues against the need to treat.

Another systematic review of vision screening tests for detection of amblyopia [2] looked for high quality studies meeting stringent criteria for evaluation of diagnostic tests. It found four studies, only one of which addressed traditional vision screening. In comparison with ophthalmologic examination the test had a sensitivity of about 12% and a specificity of 99% in a population with a prevalence of 6%.

Randomised trials of treatment [2] were small in both number and size. None compared patching with no patching


It could hardly be claimed, even by the most enthusiastic supporter of pre-school vision testing, that there is any significant evidence that screening is effective. Clearly there are many more aspects than Bandolier has time for here, but again, the evidence in these other areas is slight.


  1. SK Snowden, SL Stewart-Brown. Preschool vision screening: results of a systematic review. CRD Report 9, NHS Centre for Reviews and Dissemination, April 1997.
  2. AR Kemper, PA Margolis, SM Downs, WC Bordley. A systematic review of vision screening tests for the detection of amblyopia. Pediatrics 1999 104: 1220-1222.
previous or next story in this issue