Reply

Abstract

In life as in science, coincidences are possible—in this case the almost simultaneous publication of Dr. Hoffer et al.’s paper on protocols for studies of IOL accuracy1 and the acceptance of our study on the comparison of visual recovery between femtosecond laser–assisted cataract surgery and standard phacoemulsification. Thus, the latter had no knowledge of the former’s recommendations—which in this case does not matter much since our study was based on a different and, we strongly believe, a clinically valid and important perspective. Without invoking the old adage of comparing apples with oranges, there is no use denying that we are talking about 2 different objectives in performing clinical trials. One is the optimization of A constants. The other—pursued by us as well as by numerous authors evaluating different therapeutic options—is the clinical outcome, in this case the prediction error. Optimizing the A constant was not the topic of our study; the authors were using an optimized A constant published by Haigis et al.A (for exactly the type of IOL that was implanted in this study’s patients). It should be emphasized that it was a prospective trial with a significant number of cases; 100 eyes had femtosecond laser–assisted cataract surgery and 100 fellow eyes had conventional phacoemulsification. Also, the statistical calculation was diligently planned and based on previous clinical comparisons including capsular bag shrinkage.2 In following the recommendations of our different societies,3 the evaluation of the clinical data and the presentation of our results were in strict accordance with the guidelines of renowned journals such as the Journal of Cataract & Refractive Surgery and the Journal of Refractive Surgery. It is a well-established method and the basis for comparing the outcomes of different studies in a clinically comprehensible way. It is our pleasure to submit the data demanded by the authors of the letter. Interestingly, there were clinically irrelevant and negligible differences (Figures 1 and 2). We also strongly encourage this (or any other) group to submit proposals on their favored approach and the preferred statistical analysis to those committees of our community who are responsible for updating the guidelines to keep up with the sometimes rapid, but always fascinating, developments in cataract and refractive surgery. The prospective randomized intraindividual comparative trial we contributed comes naturally with a high level of evidence. Probably some of our readers will follow the conventional wisdom of not fixing things that are not broken.Figure 1: Spherical equivalent refractive accuracy in the femtosecond laser–assisted cataract surgery (LCS) group and in the standard group 6 months postoperatively (Datagraph version 4.20). Six months postoperatively, 71% and 100% (n = 196 eyes) in the standard group were within ±0.5 D and ±1.0 D, respectively, of their intended refractive outcome and 92% and 100% in the femtosecond laser–assisted cataract surgery group were within ±0.5 D and ±1.0 D, respectively, of the target refraction.Figure 2: Spherical equivalent refractive accuracy in the femtosecond laser–assisted cataract surgery (LCS) group and in the standard group 6 months postoperatively based on median absolute error. At 6 months, the median absolute error in both study groups showed a mean of 0.33 D ± 0.27 (SD) in the standard group and 0.23 ± 0.17 D in the femtosecond laser–assisted cataract surgery group (P = .003). The median was 0.28 D and 0.19 D, respectively. Seventy-four percent and 100% in the standard group were within ±0.5 D and ±1.0 D, respectively, of their intended refractive outcome, and 90% and 100% in the femtosecond laser–assisted cataract surgery group were within ±0.5 D and ±1.0 D, respectively, of the target refraction.