Predictability of intraocular lens power calculation after small-incision lenticule extraction for myopia

Purpose: To evaluate and compare the predictability of intraocular lens (IOL) power calculation after small-incision lenticule extraction (SMILE) for myopia and myopic astigmatism. Setting: Department of Ophthalmology, Philipps University of Marburg, Marburg, Germany. Design: Retrospective comparative case series. Methods: Preoperative evaluation included optical biometry using IOLMaster 500 and corneal tomography using Pentacam HR. The corneal tomography measurements were repeated at 3 months postoperatively. The change of spherical equivalent due to SMILE was calculated by the manifest refraction at corneal plane (SMILE-Dif). A theoretical model, involving the virtual implantation of the same IOL before and after SMILE, was used, and the IOL power calculations were performed using ray tracing (OKULIX, version 9.06) and third- (Hoffer Q, Holladay 1, and SRK/T) and fourth-generation (Haigis-L and Haigis) formulas. The difference between the IOL-induced refractive error at corneal plane before and after SMILE (IOL-Dif) was compared with SMILE-Dif. The prediction error (PE) was calculated as the difference between SMILE-Dif–IOL-Dif. Results: The study included 204 eyes that underwent SMILE. The PE with ray tracing was −0.06 ± 0.40 diopter (D); Haigis-L, −0.39 ± 0.62 D; Haigis, 0.70 ± 0.48 D; Hoffer Q, 0.84 ± 0.47 D; Holladay 1, 1.21 ± 0.51 D; and SRK/T, 1.46 ± 0.54 D. The PE with ray tracing was significantly smaller compared with that of all formulas ( P ≤ .001). The PE variance with ray tracing was σ 2 = 0.159, being significantly more homogenous compared with that of all formulas ( P ≤ .011, F ≥ 6.549). Ray tracing resulted in an absolute PE of 0.5 D or lesser in 81.9% of the cases, followed by Haigis-L (53.4%), Haigis (35.3%), Hoffer Q (25.5%), Holladay 1 (6.4%), and SRK/T (2.9%) formulas. Conclusions: Ray tracing was the most accurate approach for IOL power calculation after myopic SMILE.