Referral rate for refractive amblyopia using automated vision screening in school children in Beirut, Lebanon

Abstract

Background: Amblyopia is one of the most prevalent causes of decreased vision in children and can be effectively diagnosed at an early stage through vision screening. Untreated vision impairments during childhood have lasting implications on academic achievements. The purpose of this study was to compare referral rates when applying different referral criteria including the AAPOS 2021 exam failure levels and the Arnold "medium" 2022 Instrument Referral Criteria.

Methods: Automated vision screening was conducted in four selected schools (two private and two public) in Beirut, Lebanon. Children aged 3-6 years old were targeted. The Plusoptix A12 Refractometer was utilized to perform vision screening. Referral rate was computed for the American Association of Pediatric Ophthalmology and Strabismus (AAPOS) age-based criteria applied to the Plusoptix A12 and compared to the Arthur exam failure criteria and 4 instrument referral criteria (Arnold Medium and Specific, Matta & Silbert, and Alaska Blind Child Discovery). Referral rates were also compared between public and private schools based on the AAPOS criteria.

Results: A total of 308 children were screened: 114 students from public schools and 194 students from private schools. The gender distribution in the two groups was similar (46% females in public schools and 48% in private schools); 34% of the studied population were under 4 years old, while 66% were ≥4 years. The referral rate using the AAPOS 2021 criteria was 22%. There was a significant difference in referral rates overall and across the different types of refractive error when compared with the other referral criteria. Referral rate using the ABCD criteria was similar (17.9%). Referral rates were higher when applying Arthur criteria (28.2%) and Matta & Silbert criteria (41.9%), while the Arnold Medium and Specific criteria had the lowest referral rates (11.4%, 9.1%). The most common refractive error across all criteria was astigmatism. The overall referral rate using the AAPOS 2021 criteria was 22% and this differed significantly by school type, with a rate of 36% for public schools and 14% for private schools (p < 0.001). Spherical equivalent was also higher in public schools. Among children needing referral, 40.6% were already wearing spectacles at the time of examination and this differed by school type (4.9% of children referred from public schools and 92.8% of children referred from private schools).

Conclusion: Photoscreener referral rate and the detection of the different types of refractive errors varies significantly according to the referral criteria used. Careful consideration of psychometric characteristics of referral criteria is important and the use of device-specific age-based criteria is recommended. In our cohort, the most common refractive error was astigmatism. The overall rate of referral was 22% according to the AAPOS age-based exam failure referral criteria, which differed from referral rates applying other validated instrument referral criteria in the literature. The referral rate and spherical equivalent were higher in public schools compared to private schools.