ESCRS - PP22.01 - Accuracy Of Intraocular Lens Calculation Formulas In Eyes With Short Axial Length

Accuracy Of Intraocular Lens Calculation Formulas In Eyes With Short Axial Length

Published 2023 - 41st Congress of the ESCRS

Reference: PP22.01 | DOI: 10.82333/ch1k-gx67

Authors: Gordon T Brown* 1 , Andrea Montgomery 1 , T Graham M Cormack 1 , Magdalena Edington 1

1Ophthalmology Department,Golden Jubilee National Hospital,Clydebank, Glasgow,United Kingdom

Selection of the correct intraocular lens (IOL) power is pivotal in determining the post-operative visual outcome for patients undergoing cataract surgery. Selection of the correct IOL calculation formula is particularly important for eyes of short axial length, yet debate remains regarding which formula yields the most accurate results. The purpose of this study is to compare the accuracy of different intraocular lens calculation formulas for eyes of short axial length.

Golden Jubilee National Hospital Eye Centre, Clydebank, Glasgow, Scotland.

This single centre retrospective study reviewed outcomes for 323 patients with short axial length eyes who underwent cataract surgery between November 2015 and July 2021. Short axial length was defined as eyes measuring less than 22mm. SRK/T, Barrett Universal II, Hoffer Q and Kane were selected for comparison. Primary outcome measure was the discrepancy between desired post-operative spherical equivalent (SE) and actual post-operative SE, measured at 6 weeks following surgery. Mean absolute error (MAE), standard deviation (SD), median absolute error (MedAE) and all descriptive statistics were evaluated.  Friedman Test followed by post-hoc analysis with Bonferroni correction was used to compare differences in absolute error of each formula.

Kane produced the lowest MAE at 0.50 Dioptres (D) with SD (± 0.38). Barrett Universal II, Hoffer Q and SRK/T produced MAEs of 0.60D (± 0.48), 0.64D (± 0.47) and 0.82D (± 0.57), respectively. Kane also had the lowest MedAE at 0.43D. Barrett Universal II, Hoffer Q and SRK/T produced MedAEs of 0.5D, 0.55D and 0.71D, respectively. Kane had the highest percentage of eyes with a predicted error (PE) within 1D (90%) and 0.5D (58%).  Barrett Universal II, Hoffer Q and SRK/T were less accurate with (1D/0.5D) 82%/50%, 80%/43%, and 66%/33%, respectively. Kane was most the accurate formula, with statistically significant differences found in comparison to every other formula studied (Hoffer Q: p<0.0001, Barrett: p<0.0001, SRK/T­: p<0.0001).

In eyes of short axial length (<22mm), Kane was the most accurate predictor of post-operative SE in this study.