Surgically Induced Astigmatism.

To the Editor: We read with interest the article by Ofir et al.1 In 70 eyes, the authors used three devices to evaluate surgically induced astigmatism (SIA) in eyes undergoing cataract surgery with temporal 2.4-mm clear corneal incisions. Vector analysis was used appropriately to calculate the SIA in each eye.2 However, when reporting the median and mean of the SIA in the whole group, the authors calculated the median and mean of the magnitudes of the individual SIA vectors, independent of the vector angles (Figure 2, Table 2).1 Ofir et al. reported mean SIA magnitude values of 0.50 diopters (D) with the Lenstar LS900 (Haag-Streit, Koeniz, Switzerland), 0.53 D with the IOLMaster 500 (Carl Zeiss Meditec, Dublin, CA), and 0.56 D with the Atlas topographer (Carl Zeiss Meditec). Because SIA is a vector, it has both magnitude and angular direction, and both vector components must be included when calculating the median or mean SIA in a group of cases. To illustrate this, if in 2 patients the SIA is 0.5 D at 90° in both eyes, the mean SIA is 0.5 D. However, if the SIA is 0.5 D at 90° in one eye and 0.5 D at 180° in the other, the mean SIA is 0, not 0.5 D. We recommend using double angle plots with aggregate analysis of astigmatism, as first described by Holladay et al.3 Such a plot would clearly show both the mean SIA and the scatter of the individual measurements. We believe that, using aggregate analysis, Ofir et al. would find much lower values for mean SIA. We recently performed this analysis in our practice of Lenstar data of 25 eyes that also received 2.4-mm clear corneal incisions. The mean SIA was 0.13 D @ 95° (Figure 1), whereas the mean SIA magnitude was 0.39 D, close to the values reported by Ofir et al.1