Validity of a layperson-administered Web-based vision screening test

Abstract

PURPOSE: To determine the sensitivity and specificity of the Web-based vision-screening test (WBT) VisionForKids.org when administered by an untrained layperson in a controlled environment. METHODS: Visual acuities were obtained by an untrained layperson using the WBT under observation and by an ophthalmic professional using the Electronic Visual Acuity (EVA) Tester. Subjects were randomized to which method was used first. Subjects were considered to have failed by the following criteria: in children <48 months, 20/50 vision; in children between 48 and 59 months, 20/40 or worse; in children ≥60 months, 20/30 or worse. The results of the vision examinations by the WBT and the EVA Tester were compared; sensitivity and specificity of the WBT and the correlation coefficient to detect normal and abnormal vision compared to EVA testing were determined. RESULTS: Visual acuities were obtained on 203 children between ages 3 and 12 years (average age, 7.4 years). Sensitivity of the WBT was 78.7% (95% CI, 66.0%-87.7%), specificity was 89.4% (95% CI, 82.9%-93.8%). The correlation coefficient of EVA testing with the WBT was 0.89 (P < 0.001). CONCLUSIONS: This layperson-administered WBT, VisionForKids.org, is valid for identifying amblyopia in a controlled environment, making it possible for cost-effective and easily accessible vision screening to be performed by laypersons.