Accuracy Of Intraocular Lens Power Calculation Using Two Optical Biometres: A New Swept-Source Optical Coherence Tomography–Based Biometer And A Optical Low-Coherence Reflectometry-Based Device

To compare the application in intraocular lens (IOL) power calculations between a new swept-source optical coherence tomography (SSOCT) biometer (ANTERION, Heidelberg Engineering Ltd, Germany) with that of optical low-coherence reflectometry (OLCR) (Lenstar LS 900, HAAG-STREIT AG, Switzerland).

Private Hospital (CUF Descobertas), Lisbon, Portugal. 

Thirty-tree eyes which had undergone cataract extraction surgery with preoperative biometry with Anterion and Lenstar-LS900 were enrolled. All patients had the same intraocular lens implanted. Optical biometry and keratometry readings obtain with both devices were used to predict the refraction of the implanted IOL with six IOL power calculation formulas: Kane, Barret, SRK/T, Holladay 1, Hoffer Q and Haigis. Postoperative manifest refraction or auto-refractor measurements were obtained 3 or more weeks post-operatively. Mean prediction errors, median absolute errors, and mean absolute errors for each device using the six formulas were calculated.

With Lenstar, the Kane and Haigis formulas produced myopic refractive prediction errors (RPE) of 0.012 diopter(D) and 0.6D, respectively; all other formulas produced hyperopic RPE. With Anterion all formulas produced myopic RPE. After adjustment of the mean RPE to zero, Haigis and HofferQ had higher median absolute errors for both devices (0.46 and 0.48 for Lenstar, 0.49 and 0.49 for Anterion, respectively); there were no statistically significant differences between all formulas. Anterion had a lower percentage of eyes within an absolute deviation from target refraction of 0.50 and 1.00D for Kane, Barret, SRK/T and Haigis formulas; Holladay 1 and Hoffer Q performed similarly between both biometers. 

Both devices performed well, obtaining a mean of 65% of eyes within an absolute deviation from target refraction of 0.50D, and 90% within 1D from target refraction for each formula. The discrepancies between their measurements resulted in statistically and clinically insignificant differences in predicted IOL.