Abstract
Purpose :
Internationally vision screening programs vary significantly with no uniform protocol or standardisation of screening age. This study aims to determine the accuracy of vision screening in two age groups, 4 year old pre-school and 6 year old school children. The impact of additional tests for detecting childhood eye conditions was also assessed.
Methods :
Data on refractive error, strabismus and amblyopia from the Sydney Paediatric Eye Disease Study and Sydney Myopia Study were used in this study. Visual acuity (VA) was measured using an electronic vision chart (single-surround HOTV) in 215 4 year olds and a HOTV LogMAR chart in 1741 6 year olds with VA less than 6/9.5 requiring referral. In the 6 year olds near VA was assessed using a near HOTV logMAR chart at 40 cms. All children had a comprehensive ocular examination including cycloplegic (cyclopentolate 1%, tropicamide 1%) auto-refraction (Canon RK-F1). Spherical equivalent refraction (1/2 cyl + sphere) was calculated and significant myopia was classified as ≤-1.00D, hyperopia ≥+3.00D and astigmatism ≥1.00D. Amblyopia was defined as VA <6/12 with a difference of 2 lines in VA between the eyes and anisometropia ≥1.00D difference.
Results :
Using the 6/9.5 VA cut-off, sensitivity and specificity for the detection of the targeted eye conditions was 69.2% and 87.6% in 4 year olds and 56.0% and 94.9% in the 6 year olds, respectively. The eye conditions that the VA protocol failed to detect were primarily astigmatism (4 years: 56%, 6 years: 68.6%) and hyperopia (4 years: 24%, 6 years: 29.1%). No cases of amblyopia went undetected. The addition of a near VA test for the 6 year olds failed to detect any more cases of hyperopia. However, the inclusion of a cover test improved detection of strabismus in both age groups, with an overall further 29 cases (65%) identified. The addition of Lang II stereoacuity failed to detect any further cases of strabismus or amblyopia.
Conclusions :
VA screening is highly accurate in detecting amblyopia. In both age groups VA had moderate sensitivity but high specificity for detecting the targeted conditions. There is no indication that screening at age 6 improves accuracy compared to age 4. While cover test improves detection of strabismus, it may not be practical in a screening protocol. Near VA and Lang II did not improve detection of targeted conditions.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.