Factors Influencing Intraocular Pressure Changes after Laser In Situ Keratomileusis with Flaps Created by Femtosecond Laser or Mechanical Microkeratome

Abstract

The aim of this study is to describe factors that influence the measured intraocular pressure (IOP) change and to develop a predictive model after myopic laser in situ keratomileusis (LASIK) with a femtosecond (FS) laser or a microkeratome (MK). We retrospectively reviewed preoperative, intraoperative, and 12-month postoperative medical records in 2485 eyes of 1309 patients who underwent LASIK with an FS laser or an MK for myopia and myopic astigmatism. Data were extracted, such as preoperative age, sex, IOP, manifest spherical equivalent (MSE), central corneal keratometry (CCK), central corneal thickness (CCT), and intended flap thickness and postoperative IOP (postIOP) at 1, 6 and 12 months. Linear mixed model (LMM) and multivariate linear regression (MLR) method were used for data analysis. In both models, the preoperative CCT and ablation depth had significant effects on predicting IOP changes in the FS and MK groups. The intended flap thickness was a significant predictor only in the FS laser group (P < .0001 in both models). In the FS group, LMM and MLR could respectively explain 47.00% and 18.91% of the variation of postoperative IOP underestimation (R2 = 0.47 and R(2) = 0.1891). In the MK group, LMM and MLR could explain 37.79% and 19.13% of the variation of IOP underestimation (R(2) = 0.3779 and 0.1913 respectively). The best-fit model for prediction of IOP changes was the LMM in LASIK with an FS laser.

Fig 1. The mean IOP.after LASIK using an FS laser or MK.

Data are presented as mean ± standard deviation at 1 week, 1 month, 6 months and 12 months after surgery. In both methods of flap dissection, the postIOP at 1 month, 6 months and 12 months were all significantly lower than postIOP at 1 week (the * indicates P< .0001). The MK group showed greater postIOP from 1 week to 1 month than those of the FS laser group.

Fig 2. Distribution of the observed IOP changes and that predicted by LLM and MLR in LASIK with an FS laser.

The reference line indicates the ideal condition, in which the predicted IOP changes equal the observed IOP changes. The LMM (R² = 0.47) had superior prediction accuracy compared with the MLR model (R² = 0.1891).

Fig 3. Distribution of the observed IOP changes and that predicted by LLM and MLR in LASIK with an MK.

The reference line indicates the ideal condition, in which the observed IOP changes equal the predicted IOP changes. The LMM (R² = 0.3779) had superior prediction accuracy compared with the MLR model (R² = 0.1913).