Clinical accuracy of the AAPOS pediatric vision screening referral criteria

Abstract

The American Association for Pediatric Ophthalmology and Strabismus (AAPOS) referral criteria for amblyogenic risk factors are consensus criteria that were determined by thebest-available data as well as survey results of pediatric ophthalmologists. In 2003 the AAPOS Vision Screening Committee published guidelines to standardize reporting the ability of vision screening devices to detect these factors. We attempted to assess the accuracy of the AAPOS referral criteria. Billing records of one pediatric ophthalmologist were reviewed to identify all children who were seen in 2002. Records were excluded if photoscreening had not been performed at the initial visit or if photoscreening results were not available in the record. Of the remaining records, one-half were randomly selected for analysis. Cycloplegic refraction and binocular alignment were evaluated to determine whether the child would have been considered to be at risk for amblyopia on the basis of AAPOS referral critera. The sensitivity and specificity of these factors for detecting amblyopia was then determined. A total of 1,575 records were identified, of which 529 were randomly selected; 7 were excluded for incomplete data. AAPOS referral criteria would have referred 266 patients, of whom 255 had amblyopia and 11 did not; of the 256 patients who would not have been referred, 46 had amblyopia and 210 did not. In this population, the AAPOS referral criteria would have had an 85% sensitivity, 95% specificity, a 5% false-positive rate and a 15% false-negative rate for detecting amblyopia. Application of the AAPOS referral criteria resulted in underreporting of amblyopia in this study. We propose modifications that may result in increased sensitivity and a lower false-negative rate.