Purchase this article with an account.
AM Hulstein, KS Reuter, PP Schmidt; Performance of Autorefractors as Screening Tools in Young Children . Invest. Ophthalmol. Vis. Sci. 2002;43(13):2671.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The most frequent cause of amblyopia in early childhood is significant refractive error. Purpose: The performance of noncycloplegic retinoscopy and three handheld autorefractors is evaluated in screening for detecting hyperopia, astigmatism, and anisometropia - the most prevalent amblyogenic factors in young children. Methods: Forty-two young children (mean age: 4.7 years +/- 0.9 ) had refractive error measured in a pre-established order using cycloplegic retinoscopy, noncycloplegic retinoscopy, and three autorefractors. Nikon Retinomax K-plus and two Welch Allyn (WA) instruments [the Prototype and the Sure Sight (SS) 1.22] were completed by two testers masked to the results of retinoscopy, and the other autorefractor. Noncycloplegic and cycloplegic retinoscopy (1% cyclopentolate) were conducted separately by a single tester masked to the results of autorefractors. Cycloplegic retinoscopy data served as the gold standard measure of refractive error. Receiver operating characteristic (ROC) curves were generated for refracting methods using a range of magnitudes for hyperopia, anisometropia, and astigmatism as referral criteria. Results: Preliminary results for referral criteria of hyperopia ≷ +2.50 D, anisometropia ≷ 1 D, astigmatism ≷ 1.50 DC show that the area under the ROC curve varied for the refracting methods: hyperopia (0.75-0.96), anisometropia (0.54-0.70), and astigmatism (0.96-0.99). When sensitivity and specificity values were each maximized, the ranges reported (sensitivity, specificity) for each refractive category were: hyperopia ( 0.75-0.92, 0.62-0.93), anisometropia (0.50-0.75, 0.58-0.72), and astigmatism ( 0.88-1.0, 0.88-0.94). Further, the cutoff criteria varied for each refracting method: hyperopia (+1.00- +2.50 D), anisometropia (+0.25-2.50 D) and astigmatism (+1.50 DC) Conclusion: ROC curves show that all four refracting screening methods are weakest for detecting anisometropia and are strongest for detecting astigmatism.
This PDF is available to Subscribers Only