Predicting Suction-Induced Centration Shift During Hyperopic Smile From Preoperative Vertex-Based Geometry

Published 2026 - 44th Congress of the ESCRS

Reference: FP27.15 | Type: Free paper | DOI: 10.82333/4e9g-7t21

Authors: Ms Julia Reinstein* ¹ , Prof. Dan Reinstein ² , Dr Timothy Archer ¹ , Mr Joseph Potter ¹ , Ms Ruchi Gupta ¹

¹London Vision Clinic,London,United Kingdom, ²London Vision Clinic,London,United Kingdom;Columbia University Medical Center,New York,United States;Sorbonne Université,Paris,France;Ulster University,Biomedical Science Research Institute,Coleraine,United Kingdom

Purpose

To develop a predictive model for intraoperative shift-on-suction (SOS) during hyperopic SMILE based on preoperative ocular parameters, including angle kappa and vertex-to-limbus distance (VTLD).

Setting

Private Practice Academic Unit Reinstein Vision, London Vision Clinic, EuroEyes Group, UK

Methods

Retrospective analysis of 86 eyes undergoing hyperopic SMILE. Vertex-to-limbus distance (VTLD), white-to-white (WTW), corneal thickness (CCT), keratometry, and angle kappa (AK) were measured (MS-39). Surgical videos were reviewed to determine X and Y treatment displacement relative to the pupil centre before and after suction using the contact glass diameter as a spatial reference. Shift-on-suction (SOS) was defined as the vector difference between coordinates. Multivariate linear regression model predicted post-suction X and Y positions and were compared with measured displacement. SOS magnitude was calculated from predicted X and Y. Model performance was evaluated using leave-one-out validation with RMSE, MAE, R², and threshold accuracy.

Results

Angle kappa was the only significant predictor of post-suction displacement (AKx β=0.80, P<0.001; AKy β=0.79, P<0.001). VTLD, WTW, CCT and keratometry were not significant predictors (all P>0.05). For the prediction models, RMSE was 0.145 mm for X, 0.098 mm for Y, and 0.139 mm for SOS. MAE was 0.104 mm for X, 0.077 mm for Y, and 0.096 mm for SOS. R² was 0.81 for X and 0.53 for Y. Prediction accuracy for SOS was within 0.10 mm in 65% of eyes, 0.15 mm in 84%, 0.20 mm in 90%, 0.25 mm in 95%, and 0.30 mm in 97%.

Conclusions

Shift-on-suction is a measurable, directional centration event during hyperopic SMILE. Preoperative angle kappa predicts post-suction displacement and SOS magnitude, with internally validated accuracy for clinically significant shifts. As SMILE is centred on the corneal vertex, visual-axis offset may provide a practical parameter for anticipating suction-induced centration changes during docking.