Abstract
Although a good case for preschool screening for vision defects can be made there is very little evidence that existing programmes are effective in practice. A comparative trial of three different methods of preschool vision screening is described. Some 7000 children initially aged 5 months (younger cohorts) and 30 months (older cohorts) in three matched areas entered the trial during 1987. During 18 months of follow up new visual and ocular defects among these children were ascertained through ophthalmology outpatients and from optician records. Screening at 35 months by an orthoptist based in the community is superior to conventional health visitor surveillance at 30 months and to an agreed programme of primary care screening for squint at 30-36 months as judged by screening sensitivity (100% v 50% v 50%) and the incidence of treated target conditions (17 v 3 v 5 per 1000 person years). A notable feature in the area served by the orthoptist is that 13 children received treatment for straight eyed visual acuity loss from among 1000 children whereas there were no such cases among 2500 in the comparison areas. In the younger cohorts (that is, screening at age 5-9 months) all three programmes showed equally poor results, only one of the eight treated target conditions arising from all 3500 younger children being screen detected.