To log in, click the teal "Login" button in the upper right-hand corner of this page. If you are logged in but still do not have access, please check your 2026 Annual Meeting registration.
Moderator
Nir Sorkin, MD
Panelists
Alexander Knezevic, MD; Li Wang, MD, PhD
Viewing Papers
Expand a paper title to the right to view the paper abstract and authors. Use the video link to jump to that poster in the session.
Presenting Author
Chhavi Saini, MBBS, MPH
Co-Authors
Eric Wang (BA), Kasem Seresirikachorn (MD), Michael Bednar (BSc), Nandini Venkateswaran (MD), Daniel Vu (MD)
Purpose
To assess the frequency and direction of refractive prediction errors (RPE) across 11 intraocular lens (IOL) formulas in eyes with axial length <22 mm and determine whether myopic or hyperopic surprises predominate. Emphasis was placed on very short (<21 mm) eyes, given their higher risk of postoperative refractive inaccuracy.
Methods
This retrospective analysis included 208 eyes with axial length <22 mm implanted with a non-toric IOL and postoperative month-1 corrected visual acuity ?20/40. Predicted spherical equivalent (SE) based on the IOL power implanted was subtracted from postoperative SE to determine RPE. Myopic surprise was defined as < -0.50 D or < -1.00 D; hyperopic surprise as > +0.50 D or > +1.00 D. Outcomes included mean prediction error (PE), mean absolute error (MAE), proportion of eyes with RPE within ±0.25-1.00 D, and directionality of refractive surprises. Chi-square and exact binomial tests compared proportion of myopic vs hyperopic surprises, including subgroup analysis of <21 mm versus 21-22 mm eyes.
Results
Across formulas, MAE ranged from 0.50-0.61 D. Cooke K6 (+0.09 D) and SRK/T (+0.14 D) were the only formulas with hyperopic mean PE, while Haigis had the largest bias (-2.72 D overall). The proportion of RPE within ±0.50 D ranged from 44-64% for most formulas (76-90% within ±1.00 D), however, Haigis had the lowest accuracy vs all (47% within ±1.0 D, p<0.001). At the ±0.50 D threshold, Cooke K6, Hill-RBF and Pearl DGS gave balanced myopic and hyperopic distributions, while Barrett, EVO, Kane, Hoffer QST, Holladay 2, Hoffer Q, and Haigis had significantly more myopic surprises. Only Barrett formula showed greater myopic skew in <21 mm versus 21-22 mm eyes (p=0.01) while others were balanced.
Conclusion
Most IOL formulas demonstrated a myopic bias, while Cooke K6, Hill-RBF and Pearl DGS were balanced. Haigis showed the largest error and lowest accuracy. Only Barrett demonstrated significantly greater myopic skew in very short (<21 mm) eyes, underscoring the need for formula-specific caution in short eyes.
Presenting Author
Xuesen Cheng, MD, PhD
Co-Authors
Francisco Mendes (MD), Jay Jaber (None), Allison Chen (MD, MPH), Mitchell Weikert (MD, MS), Li Wang (MD, PhD), Douglas Koch (MD), Kusona Fortingo (BSc)
Purpose
To compare the accuracy of IOL power calculations using the Barrett Rx and MR-biometry method in eyes had prior corneal refractive surgery (CRS) with IOL exchange.
Methods
Consecutive cases underwent IOL exchange with a prior CRS from 2010 to 2024 were included, excluding eyes with IOL dislocation, unknown lens model or power in the eye, no manifest refraction at 3 weeks-6 months following exchange and DCVA ?20/40 in that interval. IOL power calculations were performed with two methods: 1) MR-biometry, incorporating data from Barrett True K, EVO and Hoffer QST; 2) the Barrett Rx formula. Refractive prediction errors were defined as the actual postoperative manifest refraction minus the predicted refraction using these two methods. Accuracy of these two methods was compared using the WHWK (Wilcox-Holladay-Wang-Koch) Statistical Software with R.
Results
Sixty-six eyes (66 patients) were included: 44 eyes had IOLs implanted fully in the bag, and 22 had sulcus-placed IOL with posterior optic capture. There were no significant differences in standard deviations (SD) or root mean square absolute errors (RMSAE) for the two formulas in either implantation scenario. In eyes with in-the-bag IOL, the SD and RMSAE of the MR-Biometry and the Barrett Rx formulas in post-myopic LASIK/PRK eyes (n = 23) were 0.45 D vs 0.40 D and 0.44 D vs. 0.43 D; in post-hyperopic LASIK/PRK eyes (n = 11) were 0.40 D vs. 0.22 D and 0.39 D vs. 0.27 D; in post RK eyes (n = 10) were 0.43 D vs. 0.70 D and 0.63 D vs. 0.67 D (all P > 0.05).
Conclusion
In this limited sample, the MR-Biometry formula demonstrated comparable refractive prediction accuracy to the Barrett Rx formula in eyes with prior CRS for IOL exchange. More data will be included.
Presenting Author
Stuti Chamola, MS
Co-Authors
Rohit Shetty (FRCS), Apurva Bansod (DNB), Abhijith Roy (MS), Raghav Narasimhan (None)
Purpose
To investigate the total corneal power changes post-lenticular surgery and assess the prediction accuracy of various intraocular lens formulas.
Methods
The study includes 200 eyes that underwent lenticular surgery using ATOS, SILK, and SMILE. The refractive error and tomography (Pentacam) were measured pre- and post-surgery. The estimated Kmean (clinical history method) was compared with the measured postoperative K mean, TCRP and EKR values. Mean prediction error and percentage of eyes within ± 0.50D, ± 0.75D, and ±1D were calculated.
Results
For all surgeries, the measured postoperative K-mean correlated significantly with estimated K-mean (P<0.001), with SMILE eyes having the best correlation (R2=0.89). For all eyes, TCRP pupil ring 3 mm correlated the best with estimated postop Kmean (P<0.001). Ray tracing (83%) is the most accurate method for estimating IOL power, followed by Haigis, Barrett True K (history and no history method).
Conclusion
For cataract surgery planning, total corneal refractive power (TCRP) measured at 3 mm should be used in IOL formulas. Ray tracing gives the most accurate estimate for IOL power calculation. Surgery-induced corneal changes vary based on the surgical platform used and should be taken into consideration during IOL power calculation.
Presenting Author
Abinaya Thenappan, MD
Co-Authors
William Wiley (MD), Shamik Bafna (MD), Kathleen Jee (MD), Sakthi Kasi (BSc)
Purpose
Evaluate tolerance to residual refractive error in post-refractive surgery eyes implanted with an advanced monofocal IOL with a slightly extended depth of focus (DOF) compared to those implanted with a standard monofocal IOL.
Methods
In this real-world retrospective analysis, data for eyes with a history of laser vision correction or radial keratotomy that underwent cataract surgery with an advanced monofocal (AM, Tecnis Eyhance) or a standard monofocal (SM, Tecnis ZCB00) IOL was captured. Uncorrected visual acuity at postop month 1 was compared between the two groups for eyes with a manifest refraction spherical equivalent (MRSE) outside the range of ±0.50 D of the target refraction. Eyes targeted for monovision and those with uncorrected posterior capsular opacification were excluded.
Results
Among all post-refractive eyes, 27.8% (32/115) of those implanted with the advanced monofocal and 27.7% (18/65) of those implanted with the standard monofocal had residual refractive error (MRSE that was not within ±0.50 D) at 1 month. Among eyes with MRSE >0.50 D and ?1.00 D, 20/20 or better uncorrected distance visual acuity was achieved in 30.4% of the AM eyes vs. 21.4% of the SM eyes; 20/25 or better in 52.2% vs 50.0%; 20/30 or better in 73.9% vs 64.3%; and 20/40 or better in 91.3% of AM vs 78.6% of SM eyes. Among eyes with MRSE >1.00 D , the corresponding percentages were 22.2% vs. 25.0%; 33.3% vs 25.0%; 55.6% vs. 25.0%; and 77.8% vs 50.0%.
Conclusion
The Tecnis Eyhance advanced monofocal IOL demonstrated greater tolerance to residual refractive error than the Tecnis ZCB00 standard monofocal IOL in post-refractive surgery eyes.
Presenting Author
Matteo Piovella, MD
Co-Authors
Barbara Kusa (MD)
Purpose
To evaluate visual performances of trifocal IOLs AT LISA tri 839 MP and AT LISA tri toric 939MP trifocal IOLs (Carl Zeiss Meditec AG - Jena - Germany) in patient that experienced previous laser vision correction to correct presbyopia
Methods
Only eyes with regular corneal map were included in this study: 38 eyes of 22 patients mean age: 56.57 ± 8.76 years. Preop SE was -0.91 ± 3,31 BCDVA 20/21.40 ± 3,18. Postop were measured: distance (5m) near (40cm) and intermediate (80 cm) VA, corneal topography, contrast sensitivity and defocus curve and quality of vision Follow-up examinations were performed at day 1 2 7 30 90 180 360 and yearly
Results
At six months BCDVA was 20/20,65 ± 2,51. SE was -0,20 ± 0,45. Residual astigmatism was 0,02 ± 0,42.83% of eyes after trifocal IOLs implantation achieved postop refractive results within ± 0.75 diopters
Conclusion
Trifocal IOLs provided good visual performances also with patients that experienced laser vision correction decades ago. To be selected for surgery eyes biometry needed to be applied with no difficulties and have to demonstrate no significant differences related the perfect IOLs power also after multiple attempts
