Background
The purpose of this paper is to validate three recent non-refractive devices that target sensory and/or strabismus risk factors for amblyopia. EyeSwift® with eye tracking and PDI Check on autostereoscopic Nintendo 3DS are video games that estimate acuity, binocularity and color. The Rebion blinq assesses foveation by birefringence.
Methods
Students were consented to be screened by each device as if by a school nurse before confirmatory examination with optimized, refined refraction. Results were compared with correlation and Bland Altman plots while screening ability was compared with 2021 AAPOS Guidelines and the strabismus-amblyopia rubric proposed by David Hunter.
Results
A total of 77 ethnically diverse students, aged 11±4 (4–19) years had a 77% prevalence of 2021 risk factors. Near visual acuity, inter-eye difference and stereopsis correlated significantly (p<0.01) for the EyeSwift® (r
2
0.14, 0.06 and 0.45) and PDI Check (r
2
0.23, 0.22 and 0.32). The sensitivity/specificity to target 2021-AAPOS: Hunter-rubric for EyeSwift® was 82%/56%: 74%/62%, for PDI Check was 75/64%: 69%/69% and for blinq 75/48%: 82%/39%. Sensitivity/specificity for the seven color-deficient students was 86%/84% for EyeSwift® and 100%/81% for PDI Check. Screening time was 96±19 s for PDI Check and 375±102 s for EyeSwift but less than 25 s for blinq.
Conclusion
Reliable near visual acuity, stereo and color testing were obtained from the two binocular and touch screen devices often with minimal coaching of the student. The blinq rapidly gives an estimate of whether both eyes are steadily fixing, and the ProVersion predicted 100% which eye is deviated or amblyopic. Non-refractive validation metrics were beneficial.