Comment on: Automated and subjective refraction with monofocal, multifocal, and EDOF intraocular lenses: review.

We read with great interest the article by Bellucci et al., in which the authors described the lack of agreement between automated refraction and subjective refraction across different intraocular lens (IOL) technologies.1 This observation led to clinically relevant take-home messages which are in complete agreement with our recent review.2 Among the standout points are the systematic myopic bias of automated refraction and the accuracy of astigmatism measurements. Moreover, the authors have proposed several insightful points to alert users to potential errors. These points include the suggestion to incorporate information about the imprecision of measuring eyes with high-technology IOLs in the autorefractor manual and to describe this error in the accompanying IOL information sheet. They also suggest encouraging researchers to include automated refraction as a complement, not a replacement, for subjective refraction in every study. This advice could provide valuable information to users adopting this technology, particularly when deciding the starting point for subjective refraction. The authors also highlighted several improper practices involving the use of automated refraction for clinical decision-making. These include instances where unnecessary spectacle correction may be prescribed or, in the worst-case scenario, this misunderstanding may lead to inappropriate refractive corneal surgery. However, it seems that the authors may have used the term “visual acuity” when they intended “subjective refraction,” initiating the discussion with the sentence, “Automated refraction remains the simplest and the most effective way to start measuring visual acuity.” Although the authors did not clearly state this in the review, in our view, the key points that need emphasis from this review are, first, that automated refraction should be considered the standard starting point in subjective refraction measurement as it is more precise than aberrometry refraction, a point we previously highlighted.2 Second, and more importantly, it should never be used either for measuring corrected distance visual acuity or for making decisions in clinical practice such as adjusting a constant for IOL power calculation, prescribing spectacles, or, at the very least, for making individual decisions such as conducting a refractive corneal surgery. It should not be forgotten that neural processing and neuroadaptation, especially in cylinder prescription, must be considered when making individual clinical decisions.3,4 Moreover, subjective refraction should never be omitted, as has been the practice in some scientific publications on IOLs. Furthermore, although a systematic bias can be corrected, the limits of agreement can be increased up to 1 diopter (D) for the spherical equivalent in comparison with the common 0.5 D shown for the intraexaminer and interexaminer reproducibility of subjective refraction.2,5 Thus, the decision to replace subjective with objective refraction in making individual decisions can critically impair a patient’s visual performance. In this letter, we are not suggesting that the authors are proposing to replace subjective refraction by automated refraction. Rather, as the authors have indicated, this is a practice that continues to be observed in the field and it deserves special attention to emphasize that objective refraction is a starting point and should not replace subjective refraction.