Abstract
Purpose:
The Pediatric Vision Scanner (PVS) directly detects strabismus and amblyopia by analyzing binocular scans for retinal birefringence, which is characteristic of steady, bifoveal fixation. In a preliminary study (AAPOS 2012), the PVS demonstrated 98% sensitivity and 88% specificity. Here we report the sensitivity and specificity for a larger cohort, with confidence intervals, and analyses of screening errors.
Methods:
250 children (2-6y) were enrolled; 147 with strabismus and/or amblyopia, and103 controls. In addition to the PVS, children were tested with the SureSight™ Vision Screener and the Randot® Preschool Stereoacuity Test (RPST). Each test yielded a recommendation of “pass” or “refer.” A comprehensive pediatric ophthalmic exam served as the gold-standard.
Results:
The PVS correctly identified 144 of 147 children with strabismus and/or amblyopia, [sensitivity = 98% (95%CI: 95-100%)] and 90 of 102 control children [specificity = 88% (95%CI: 80-96%)]. The PVS missed 3 children with amblyopia and referred 12 control children. The SureSight™ (sensitivity = 63%; specificity = 69%) and RPST (sensitivity = 79%; specificity = 77%) were less accurate. The SureSight™ missed 49 and the RPST 29 children with strabismus and/or amblyopia. The SureSight™ mistakenly referred 32 and the RPST 29 control children. Using Bayesian analysis to estimate test performance in a preschool screening setting (assuming 4% prevalence), accuracies of 89%, 69%, and 77% were predicted for the PVS, SureSight™, and RPST tests, respectively.
Conclusions:
The PVS identifies children with strabismus and/or amblyopia with high sensitivity, outperforming the SureSight™ and the RPST. Preschool vision screening may be more efficient with a device that directly detects strabismus and/or amblyopia.
Keywords: 417 amblyopia •
722 strabismus •
709 screening for ambylopia and strabismus