Clinical Comparability of a Computer-Based Vision Screening Platform to Snellen Acuity in a Young Adult Population

Authors

• Meghan E. Garvey, University of South Dakota; Leidos, LLCFollow
• Lindsey Stroop, Leidos Health ServicesFollow
• Jacob Just, University of South DakotaFollow
• Brynn Palmer, University of South DakotaFollow
• Andrew Burchett, University of South Dakota; Leidos, LLCFollow
• Donald J. Kosiak Jr, University of South Dakota, Sanford School of MedicineFollow

Abstract

Background: Military readiness depends on timely access to accurate health screening tools. The use of digital technologies for visual acuity testing may reduce bottlenecks during large-scale Physical Health Assessment (PHA) events and improve the efficiency of provider workflow. This study evaluated a computer-based vision screening (CBVS) platform against the traditional Snellen chart in a non-military, college-aged sample to assess clinical comparability and test-retest reliability.

Methods: Visual acuity was assessed in 68 college-aged adults (77.9% female, 73.5% Non-Hispanic White; mean age = 20.9 years) using both a Snellen chart and the CBVS platform, with assessments performed in randomized order to minimize order bias. Scores were recorded for each eye and converted to Snellen-equivalent units. A subset (n = 14) completed a follow-up session for test-retest analysis. Paired differences in raw scores between methods and sessions were assessed using Wilcoxon signed-rank tests. Acuity scores were also categorized into functional vision groups (Normal =45), and agreement was evaluated using prevalence-adjusted bias-adjusted kappa (PABAK). Statistical analyses were conducted in Stata 19.5, with significance set at p < 0.05.   Results: CBVS acuity scores differed significantly from Snellen measurements for both eyes (Left: z = -5.79, p < .001; Right: z = -5.55, p < .001), with CBVS generally reporting worse acuity. Test-retest comparison of CBVS showed no significant difference between sessions (Left: z = -1.88, p = .08), indicating acceptable short-term reliability. Functional vision groups observed agreement between CBVS and Snellen classifications was high (Left: 94.2%; Right: 98.6%). Prevalence-adjusted bias-adjusted kappa (PABAK) values further supported strong agreement across clinically relevant vision categories (Left = .88, Right = 0.97).   Conclusion: Although statistically different, the differences were not clinically meaningful within readiness thresholds and displayed consistent performance on retesting. These results support the potential use of digital vision screening technologies in pre-screening workflows at PHA events. The sample population consisted primarily of young adults with high visual acuity. Future studies should expand to military-age populations and evaluate implementation efficiency in field settings.

Recommended Citation

Garvey ME, Stroop L, Just J, Palmer B, Burchett A, Kosiak DJ. Clinical Comparability of a Computer-Based Vision Screening Platform to Snellen Acuity in a Young Adult Population. Aesculapius. 2026 Feb 09; 7(2026):Article 3. Available from: https://red.library.usd.edu/aesculapius/vol7/iss2026/3. Free full text article.