The purpose of this study is to determine the optimal screening intervals and cost-effectiveness of population-based vision screening in preverbal children in the Netherlands.

Amblyopia (prevalence 3 - 4%)is a preventable, unilateral loss of vision in young children, in most cases caused by strabismus (squint) or unilateral hypermetropia (need of plus glasses), or both. The sensitive period in which vision loss can develop and be recovered, by covering the better eye with a patch, is up to 6 years of age. Population-based programs for child vision screening exist in Sweden, the United Kingdom, Canada, eastern European countries and the Netherlands. Measurement of visual acuity at age four (preschool) is most common. In some countries, including Canada and the Netherlands, preverbal screening of visual function in infants and young children (age 0 - 2) has been added to the screening program. In the Netherlands, it is applied during the regular visits at ages 1-2, 3-4, 6-9, 14-24 months. At the age of 36 months, monocular visual acuity is tested with a picture chart. At the age of 45 and 60 months, monocular visual acuity is measured with the Landolt-C chart. In 1996, the investigators started a follow-up study of a birth cohort in Rotterdam (RAMSES) to determine the sensitivity, specificity and effectiveness of the screening program. A diagnosis of amblyopia was made in 100 (3.4%) of 2,964 children and was caused by refractive error (42), strabismus (19), both combined (30) or deprivation (7). It was found that most cases of amblyopia were detected by vision screening with measurement of visual acuity from age 3 years onwards. Preverbal screening enabled earlier detection of strabismus amblyopia, but not at all of refractive amblyopia.

The investigators therefore propose a disinvestment study of vision screening, with omission of screening at age 6-9 and 14-24 months. First, the optimal screening intervals are calculated with a newly developed micro-simulation model for effectiveness of repeated screening, on the basis of the data obtained in the RAMSES study. The model simulation predicted that screening at age 6-9 and 14-24 months can be omitted without an appreciable loss of number of detected amblyopia cases. This will be tested in a RCT among two large youth health care organisations, ICARE (12,500 new children annually) and GGD-Amsterdam. Two large birth-cohorts will be recruited. Children born between July and December 2011 will form our control group and will be examined at 6-9, 14-24, 36 and 45 months. Children born between January and June 2012 will be our intervention group. These children will be examined only at 36 and 45 months.

Endpoint is the number of cases of amblyopia detected, weighted for visual acuity, against age at detection. The question is whether a significant difference between groups can be found in the cumulative number of cases of amblyopia detected up to the age of 4 years. The new model for optimising screening intervals will be developed further in the course of the study into a generic tool for determining optimal screening intervals in any screening program with repeated exams for young children and should ultimately allow for comparison of effectiveness of population-based screening programs for different disorders in different countries, carried out by different personnel. Finally, in the RAMSES study 23% of the children screened positively had not been referred successfully to an ophthalmologist or orthoptist; 43% of their parents had low to moderate fluency in the Dutch language. This reflects lack of utilisation among immigrants of preventive screening and a faltering transit from prevention to care. Unsuccessful referral, in relation to parental fluency in Dutch, ethnicity and SES will, hence, be monitored in the study in an exploratory fashion.

Inclusion Criteria:

All children visiting the selected screening offices and becoming 6 months of age(in a 6 months period from the start of recruiting).

Exclusion Criteria:

• previous eye surgery
• previous eye disorder