Effectiveness of the iPhone GoCheck Kids smartphone vision screener in detecting amblyopia risk factors.

BACKGROUND: The recently released GoCheck Kids iPhone photoscreening app is designed to detect amblyopia risk factors (ARFs) in young children and includes remote review of images captured by smartphone. We aimed to evaluate the system's accuracy in detecting AAPOS guidelines ARFs. METHODS: Patients 6 months through 6 years of age at Medical University of South Carolina were recruited for this prospective study. Presence of age-specific ARF was determined based on a complete ophthalmic examination and compared with the GoCheck Kids recommendations. RESULTS: A total of 244 children were included (average age, 42 months; 51% male). Sensitivity of the GoCheck Kids iPhone with remote review to detect ARFs was 90.5%; specificity, 68.1%; positive predictive value, 56.8%; negative predictive value, 94.0%. Two of the 7 false negative results had hyperopia of >4 D, 4 had astigmatism, and 1 had anisometropia. Remote review of all images improved sensitivity. CONCLUSIONS: The GoCheck Kids app had good sensitivity and adequate specificity in detecting AAPOS ARFs in our enriched cohort of young children.

Effectiveness of the GoCheck Kids Vision Screener in Detecting Amblyopia Risk Factors.

PURPOSE: The GoCheck Kids smartphone photoscreening app (Gobiquity Mobile Health, Scottsdale, Arizona, USA), introduced in 2014, is marketed to pediatricians with little published validation. We wished to evaluate the GoCheck Kids Screener for accuracy in detecting amblyopia risk factors (ARF) using 2013 American Association for Pediatric Ophthalmology and Strabismus guidelines. DESIGN: Validity assessment. METHODS: Children 6 months to 6 years of age presenting from October 2016 to August 2017 were included. Children were screened with the GoCheck preloaded Nokia Lumia 1020, software version 4.6 with image processing version R4d, prior to undergoing a comprehensive eye examination by a pediatric ophthalmologist masked to the screener results. Determination of the presence of age-specific ARF was made based upon the examination and compared with the GoCheck recommendation. RESULTS: A total of 206 children were included (average age 43 months). When compared to examination, GoCheck had a sensitivity of 76.0% and specificity of 67.2% in detecting ARF. Positive predictive value was 57.0% and negative predictive value 83.0%. The screener results of 13 children were changed from "no risk factors" to "risk factors identified" based on the GoCheck remote review process. Four images remained "not gradable" and screening was unsuccessful in 3 children. CONCLUSION: In our high-risk population, this version of the Gocheck Kids smartphone app was useful in identifying ARF in children who are often not able to cooperate with visual acuity testing. This study informs pediatricians about the efficacy of this new screener as they make decisions about how to best detect vision problems in young children.