Background  

Photorefraction can be carried out in both eyes simultaneously from a distance and is therefore suitable for examination of children. This study evaluated the accuracy of a commercially available photorefractometer (PowerRefractor, Plusoptix, Erlangen, Germany) and investigated whether the working distance relaxes the accommodation sufficiently without cycloplegia.

Methods  

Photorefractometer readings were compared to cycloplegic retinoscopy. Because of the limited working range the group of patients consisted of low and moderate ametropic eyes within a spherical power of –7.0 to +5.0 D. One hundred and ninety-two eyes from 104 patients (2–81 years) were photorefracted under cycloplegia. A subgroup of 83 eyes from 46 patients was additionally refracted without cycloplegia.

Results  

Under cycloplegia, the PowerRefractor measured the spherical equivalent slightly below that of cycloplegic retinoscopy (too much minus). The mean difference in spherical equivalent was –0.12±0.91 D (SD). The mean difference of cylindrical power was –0.17±0.73 D. The mean weighted axis difference was 0.61±0.71 D which is comparable to an axis deviation of 18° at a cylinder power of 1.00 D. Without cycloplegia, the mean difference of the spherical equivalent was –0.73±1.25 D. The mean difference of cylindrical power was –0.20±0.65 D. The mean weighted axis difference was 0.44±0.58 D which is comparable to an axis deviation of 13° at a cylinder power of 1.00 D.

Conclusions  

Without cycloplegia, the spherical equivalent of the PowerRefractor tends to be underestimated due to uncontrolled accommodation, especially in children. Cycloplegia improves the accuracy in evaluating the spherical equivalent, but decreases the accuracy of cylinder power and axis.