RESUMO
PURPOSE: To evaluate the performance of the autorefractor Retinomax K-plus2 and the photoscreener plusoptiX S08 in measuring refractive errors by comparing them with cycloplegic retinoscopy (CR) and to assess limitations associated with their use. METHODS: Cross-sectional study to compare data from CR, performed by an orthoptist, to data from Retinomax K-plus2 and plusoptiX S08 performed by a lay screener. Sensitivity and specificity for the detection of significant refractive errors were determined according to American Academy of Pediatric Ophthalmology and Strabismus criteria. RESULTS: Two hundred children were included, with a mean age of 5.2 ± 2.6 years (3 months to 11 years). Compared to CR, the plusoptiX S08 showed a mean difference of -1.13 ± 1.25 D (95% limits of agreement [LOA], -3.59 to +1.32) for spherical equivalent (SE) and -0.23 ± 0.53 D (LOA, -1.28 to +0.81) for the cylinder. Mean difference for the Retinomax K-plus2 before cycloplegia was -0.08 ± 0.58 D (LOA, -1.23 to +1.06) for SE and 0.03 ± 0.38 D (LOA, -0.72 to +0.78) for the cylinder; after cycloplegia -2.11 ± 1.64 D (LOA, -5.33 to +1.10) for SE and -0.06 ± 0.47 D (LOA, -0.98 to +0.86) for the cylinder. Sensitivity for detecting hyperopia >3.5 D with the plusoptiX S08 was 33.3%, the Retinomax before cycloplegia 31.0% and after cycloplegia 84.6% and high for detecting myopia, astigmatism, and anisometropia. CONCLUSIONS: Retinomax K-plus2 and plusoptiX S08 have high sensitivity for the detection of myopia, astigmatism, and anisometropia compared to cycloplegic retinoscopy; however, when used without cycloplegia, hyperopia is underestimated.