Comparison of 2Win and plusoptiX A12R refractometers with Retinomax handheld autorefractor keratometer

To test the accuracy and validity of the 2Win and the plusoptiX A12R refractometers in detecting amblyopia risk factors.

Children were screened using both devices, using two sets of referral criteria each, and underwent complete ophthalmic examination, including cycloplegic refraction. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. Median values for the pairs of refractometers were compared using the Wilcoxon signed-rank sum test for sphere, cylinder, and power vectors J0 and J45 for axis (both eyes).

A total of 284 eyes of 142 children (mean age, 37.9 ± 19.8 months) were included. Comparison of mean cycloplegic and manifest refractometer measurements provided statistically significant differences in both eyes. For sphere, the means were lower and for cylinder, higher for both devices, and both correlated well with the gold standard for astigmatic power vectors J0 and J45. Using referral criteria 1, the sensitivity, specificity, PPV and NPV, and inconclusive results were 67.4%, 83.7%, 87.9%, 59.4%, and 4.9%, respectively, for the 2Win, with sensitivity of 13% in the hyperopia group; 73.1%, 95.9%, 96.6%, 69.1%, and 10.6%, respectively, for the plusoptiX A12R, with a sensitivity of 33.3% in the hyperopia group. Using criteria 2, the values were 98.8%, 38.8%, 73.9%, and 95.0% (2Win) and 94.9%, 65.3%, 81.3%, and 88.9% (plusoptiX A12R).

In manifest conditions, the accuracy of the 2Win and plusoptiX A12R refractometers is low in hyperopia and astigmatism, but the devices are well correlated with each other, and both have high specificity; sensitivity is low in hyperopia, resulting in the underestimation of hyperopic refractive error. The optional Plusoptix sensitive referral criteria seems to be appropriate for the A12R. The 2Win provided fewer inconclusive results and was preferred for use with younger and developmentally delayed children.