Abstract
Purpose: :
The purpose of this study is to evaluate the ability of the SureSight Vision Screener and Retinomax Autorefractor to detect specific refractive errors (hyperopia, myopia, astigmatism, and anisometropia).
Methods: :
As part of the VIP Study, 1446 Head Start preschool children had refractive error measured by the SureSight Vision Screener (child mode) and Retinomax Autorefractor, followed on a later day by cycloplegic retinoscopy by a masked pediatric eyecare professional. Significant refractive errors on cycloplegic retinoscopy were defined as: hyperopia >+3.25D in any meridian; myopia > -2.00D in any meridian; astigmatism > 1.50D between principal meridians; and anisometropia > 1.00D interocular difference in hyperopia (most hyperopic meridian), or > 3.00D interocular difference in myopia (most myopic meridian), or > 1.50D interocular difference in astigmatism. The ability of the SureSight and Retinomax measurements to identify each refractive error was summarized by the area under the receiver operating characteristic (ROC) curve using all possible cutpoints for defining failure. Detection of refractive error type was based upon the child’s worse eye and using the following SureSight and Retinomax
Results: :
most positive meridian for hyperopia, most negative meridian for myopia, cylinder for astigmatism and maximum of intereye difference in the sphere and cylinder for anisometropia.Results: The criterion for significant refractive error was met by 168 children for hyperopia, 25 for myopia, 156 for astigmatism and 94 for anisometropia. Area under the ROC curve was greatest for myopia and astigmatism (SureSight:0.95; Retinomax:0.96), followed by hyperopia (SureSight:0.83; Retinomax:0.88), and anisometropia (SureSight:0.78; Retinomax:0.67). Comparison of area under ROC curves revealed no significant difference between instruments for myopia and astigmatism, but significantly better performance by Retinomax for hyperopia (p = 0.03) and significantly better performance by SureSight for anisometropia (p = 0.001).
Conclusions: :
Detection of significant refractive error with SureSight Vision Screener and Retinomax Autorefractor was very good for astigmatism and myopia, good for hyperopia, and fair for anisometropia. Neither machine was uniformly better than the other for all four targeted conditions.
Clinical Trial: :
www.clinicaltrials.gov NCT00038753