Comparison Of Refractive Prediction Error After Intrascleral Intraocular Lens Fixation In Eyes With Long Axial Length
Published 2026 - 44th Congress of the ESCRS
Reference: PP03.13 | Type: Free paper | DOI: 10.82333/0aes-ec94
Authors: Dr Masaki Suzue* 1 , Dr So Goto 1 , Dr Susumu Sakimoto 2 , Prof. Dr Takatoshi Maeno 1 , Prof. Dr Kohji Nishida 2
1Department of Ophthalmology, Graduate School of Medicine,The University of Osaka,Suita,Japan, 2Department of Ophthalmology, Graduate School of Medicine,The University of Osaka,Suita,Japan;Integrated Frontier Research for Medical Science Division, Institute for Open and Transdisciplinary Research Initiatives (OTRI),The University of Osaka,Suita,Japan
Purpose
Since its introduction by Yamane et al. in 2017, flanged intrascleral intraocular lens (IOL) fixation using the double-needle technique (SF-IOL) has gained widespread adoption due to its shorter operative time and favorable safety profile. Although standard IOL power calculation formulas are often applied to SF-IOL, their accuracy in myopic eyes remains unclear. The aim of this study was to compare postoperative prediction error (PE) between eyes with long and normal axial length (AL) undergoing SF-IOL.
Setting
This retrospective, single-center clinical study was conducted at The University of Osaka Hospital, Osaka, Japan. We reviewed the medical records of all consecutive patients who underwent SF-IOL between January 2023 and April 2024. In all cases, surgeries were performed by seven retinal specialists using SF-IOL. All patients underwent Pars Plana Vitrectomy (PPV). A 7-mm diameter IOL was implanted in every patient.
Methods
Predicted refractive outcomes were calculated using four formulas: the Barrett Universal Ⅱ (BU Ⅱ), Hoffer Q, Holladay 1, and SRK/T formulas. Eyes were categorized into two distinct groups based on preoperative AL: the normal AL group (N group, AL < 26.0 mm) and the long AL group (L group, AL ≥ 26.0 mm). PE was defined as the difference between the postoperative spherical equivalent refraction at 3 months and the predicted refraction calculated by each formula. Refractive prediction accuracy was compared between the N and L groups. The standard deviation (SD), the slope of the simple liner regression between AL and PE, and the Formula Performance Index (FPI) were assessed for each formula.
Results
A total of 53 eyes from 50 patients were included in this study, 35 eyes in the N group and 18 eyes in the L group. Mean AL was 24.6±0.8 mm in N group and 27.1±1.0 mm in L group. Mean PE (N and L group) was −0.55±0.72 D and −0.64±0.62 D with BU Ⅱ (P=0.52), −0.68±0.77 D and −0.45±0.62 D with Hoffer Q (P=0.60), −0.78±0.74 D and −0.31±0.72 D with Holladay 1 (P=0.043), and −0.74±0.72 D and −0.62±0.72 D with SRK/T (P=0.32). In L group, Holladay 1 showed PE nearest zero; however, demonstrated steep AL–PE regression slope and low FPI. BU II yielded the lowest SD and highest FPI.
Conclusions
The Barrett Universal Ⅱ formula demonstrated the highest FPI in eyes with long AL undergoing SF-IOL, along with the most favorable values for SD and AL–PE slope. The Holladay 1 formula demonstrated the lowest mean prediction error in long AL eyes, because its hyperopic bias in long eyes offset the myopic shift observed in this study.