A traditional autorefractor outdoes a newer retinal birefringence scanner for detecting eye disorders in children.A diagnostic accuracy study shows greater overall diagnostic sensitivity with a traditional autorefraction device than with a newer, distinct photoscreening device—a retinal birefringence scanner (RBS)—for detecting amblyopia and refractive errors in children.
Among participants who screened positive for a reference standard diagnosis, the traditional autorefractor was more sensitive than the RBS for amblyopia (90 percent vs 62 percent; p=0.03) and visually significant refractive error (VSRE; 95 percent vs 54 percent; p<0.001) but not for strabismus (55 percent vs 61 percent; p=0.69).
According to the investigators, the discrepancy may have been driven by the difficulty of detecting micro-strabismus or small-angle deviations using optical reflex-based measurements. “We used a relatively inclusive definition of strabismus (≥5 prism dioptres). [Hence], it is possible that this threshold could have reduced the apparent sensitivity of the devices for detecting amblyogenic strabismus.”
In children with no vision disorder, specificity was slightly lower with the traditional autorefractor than with RBS (69 percent vs 74 percent; p=0.70). [JAMA Ophthalmol 2026;doi:10.1001/jamaophthalmol.2026.1531]
The traditional autorefractor had greater diagnostic accuracy than the RBS, with a greater AUROC* for amblyopia (0.96 vs 0.67; p<0.001), strabismus (0.76 vs 0.69; p=0.17), VSRE (0.92 vs 0.66; p<0.001), and the composite of amblyopia, strabismus, or VSRE (0.80 vs 0.66; p=0.01).
Screening technologies need refinement
“Our findings provide insight into the comparative strengths and limitations of these two screening devices for detecting eye disease in children,” the investigators noted. “Understanding the differences in sensitivity and specificity for different devices can allow efficient evidence-based screening programmes to improve early detection of childhood eye disease while reducing unnecessary referrals.”
The study enrolled 195 children referred to a specialized paediatric ophthalmology clinic at the University of California, San Francisco, for suspected vision problems. Each child underwent screening with a traditional autorefractor (Spot Vision Screener) and an RBS (Blinq vision scanner), followed by a comprehensive eye exam. A total of 139 children (mean age 49.7 months, 54 percent female) completed both photoscreening exams.
The two devices have varying mechanisms. The traditional device uses an autorefractor to detect amblyopia risk factors (eg, refractive error, eye misalignment), whereas the RBS aims to directly identify amblyopia by evaluating abnormal binocular foveal fixation. [Clin Ophthalmol 2016;10:1365-1373; Pediatrics 2017;140:e20163745; Invest Ophthalmol Vis Sci 2011;52:5043-5048]
Fifteen percent of participants had amblyopia (52.4 percent unilateral, 47.6 percent bilateral), and 30 percent had VSRE, including astigmatism, hyperopia, anisometropia, and myopia (16.9, 10.3, 8.9, and 4.4 percent, respectively).
Approximately 37 percent had strabismus (25.2 percent distance strabismus, 32.4 percent near strabismus). The most common strabismus subtype was exotropia (13.7–15.8 percent); the least common was vertical strabismus (<3 percent).
However, the study lacked sufficient power to evaluate specific subtypes of strabismus or amblyopia. The multiple screenings may have also led to fatigue or inattention, and the results may not be generalizable to primary care or community screening settings, as the study was conducted in a clinic serving kids with prior failed vision screening.
“[Nonetheless, the] findings suggest that the traditional autorefractor may be more reliable for comprehensive paediatric vision screening,” the investigators said.
“Our findings underscore the need for ongoing refinement of screening technologies and standardized, transparent reporting to guide clinical implementation and optimize early detection of treatable vision disorders in children,” they added.