Diagnostic accuracy of the Parr vision test, single crowded Lea symbols and Spot vision screener for vision screening of preschool children aged 4–5 years in Aotearoa/New Zealand

Purpose Preschool children in New Zealand undergo vision screening to detect amblyopia at 4-5 years of age. The current test, the Parr vision test, does not meet international visual acuity chart guidelines and has not been validated against other commonly used paediatric vision tests. New Zealand vision screening protocols are also not targeted for detecting other eye conditions such as uncorrected refractive error, which may affect school performance. We compared the Parr vision test with the single crowded Lea symbols and the Spot vision screener for detecting ocular pathology, refractive error and amblyopic risk factors in preschool children. Methods A cross-sectional diagnostic accuracy study recruited children aged 4-5 years via convenience sampling from the University of Auckland Optometry Clinic and through primary schools in Auckland, New Zealand. Participants received vision screening with the three different instruments administered by a lay screener. Comprehensive eye examinations were completed by a paediatric optometrist to determine the presence of vision disorders. Results Of 197 children who received a comprehensive eye examination, 14 (7.1%) had amblyopic risk factors and 43 (21.8%) had significant refractive error (15.7% with astigmatism, 9.1% with hyperopia). The sensitivity for detecting any ocular condition did not differ significantly between the tests (50.0% for Parr, 43.5% for Lea, 42.5% for Spot). Specificity was significantly lower for the Parr vision test (80.8%) than for the Lea symbols (93.4%) and Spot vision screener (98.0%). Adding the Spot vision screener to measurements of visual acuity significantly improved sensitivity in detecting any ocular condition with the Parr vision test (67.5% for Parr/Spot vs 50% for Parr alone), but not with the Lea symbols (52.5% for Lea/Spot vs 43.5% for Lea alone). Conclusion The sensitivity of the Parr vision test for detecting ocular conditions in preschool children does not vary significantly from that achieved by the Lea symbols or the Spot vision screener. However, current New Zealand vision screening protocols could be improved by expanding the target conditions to include significant refractive error and incorporating the use of the Spot vision screener to increase the accuracy with which children with refractive error are identified. Future research should include longitudinal studies to determine the effect of preschool vision screening on later ocular and academic outcomes.