Abstract
Purpose :
To assess the accuracy of visual acuity screening programs involving ‘all class teachers’(ACTs)(intervention) compared with standard program involving a limited number of ‘selected teachers’(STs)(control) (1 teacher per 250 children selected by school head) in southern Indian schools.
Methods :
This was a prospective, non-randomized cluster-controlled trial of alternate school teacher, visual acuity screening models. Three schools from in and around Pondicherry, India, were selected for intervention and 3 were matched in terms of age and sex of children, type of school and distance from base hospital as controls. Teachers were trained to identify students between the age of 6 and 17 with visual acuity of 20/30 or worse in either eye or obvious ocular abnormalities, such as squint, and to refer. Visual acuity (VA) was measured by teachers using the “E Chart” for one eye and the Snellen Number Chart for the second eye at a distance of 6 meters. An ophthalmic team, including an ophthalmologist, visited schools to examine all children screened, regardless of screening results. Sensitivity and specificity, referral compliance and cost of case finding were compared.
Results :
ACTs (n=82 teachers) screened 3,574 and STs (n=20 teachers) screened 3,562 children. Significantly more referred children from ACTs reached the base hospital for further investigations within three months compared to STs [(78.6% vs 18.4 %) (p<0.05)]. Both sensitivity and specificity of ACTs model were higher than STs (sensitivity 93.6 vs 67.2%, and specificity 94.3% vs 85.6%). Cost of screening per child with visual impairment or actual ocular pathology was lower for ACTs (US$4.40) compared to STs (US$7.03).
Conclusions :
A school vision screening program involving ACTs had greater sensitivity, specificity and four times the referral uptake rate of STs at about two-thirds of the screening cost/child. Savings with ACT were largely due to reductions in un-necessary referrals.
This is a 2020 ARVO Annual Meeting abstract.