ABSTRACT
INTRODUCTION: The prospective Vision in Preschoolers (VIP) study evaluated 11 methods of screening and proposed referral criteria for the Welch Allyn SureSight(trade mark) Vision Screener with 90% and 94% specificity. The SureSight had a higher sensitivity than most other screening techniques when these criteria were applied. We evaluated the usefulness of these criteria in a field study of healthy preschool children. METHODS: The SureSight software was altered to recommend referral using the VIP referral criteria with 90% specificity. Lions Club volunteers screened preschool children throughout Tennessee. Referred children underwent comprehensive eye examinations with cycloplegic refraction. Examination failure criteria were based upon published standards. Reanalysis using the 94% specificity criteria was then performed. Outcomes included referral rate and positive predictive value. RESULTS: The SureSight was used to screen 4,733 children, and screening was successful in 99.7% of children. The referral rate using the 90% specificity criteria was 12.2%. Most children (73%) were referred for suspected astigmatism. The positive predictive value was 30%. Using the 94% specificity criteria from the VIP study decreased the referral rate to 7.9% and substantially decreased over referral for suspected astigmatism; however, several anisometropes went undetected. Higher specificity was achieved by raising astigmatism referral criteria to 2.2 diopters while leaving the anisometropia criteria unchanged. CONCLUSIONS: The SureSight can be used successfully for preschool screening in the field provided that criteria with high specificity are incorporated into the instrument's software program. Higher rates of positive predictive value can be achieved without jeopardizing sensitivity by raising astigmatism referral criteria to 2.2 diopters.
