Moderator
Morgan Micheletti, MD, ABO, FACS
Panelists
Li Wang, MD, PhD
Mohamed Abou Shousha, MD, ABO
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
Damien Gatinel, MD, PhD
Co-Authors
Guillaume Debellemani�re MD, Alain Saad MD, Avi Wallerstein FRCSC, MD, Jacques Malet PhD
Purpose
This study investigates the effect of prioritizing the reduction of the standard deviation (SD) of prediction errors as part of the optimization process, before adjusting the mean prediction error (PE), on the precision and accuracy of various IOL power calculation formulas.
Methods
A retrospective analysis was conducted on 4,885 eyes from 2,611 patients, all of whom received the same IOL model. Four commonly used IOL power calculation formulas�SRK/T, Holladay 1, Haigis, and Hoffer Q�were compared. We introduced new constants designed to minimize the SD of PE (new_const) and compared them to traditionally optimized constants (classic_const) using a heteroscedastic statistical approach. Precision improvements were validated using a secondary dataset comprising 262 eyes from 132 patients.
Results
Significant reductions in mean absolute error (MAE) were observed across both training and test datasets for the Hoffer Q, Holladay, and Haigis formulas, indicating enhanced accuracy. The optimized constants led to notable reductions in SD for the Haigis formula (from 0.3255 to 0.3153) and the Hoffer Q formula (from 0.3521 to 0.3387). These optimizations also increased the percentage of eyes with a PE within �0.25 D. While the SRK/T formula showed a minor SD improvement (from 0.3596 to 0.3585), the Holladay 1 formula exhibited minimal change with no significant improvement. In the test dataset, significant SD reductions were observed for both the Haigis and Hoffer Q formulas.
Conclusion
Prioritizing the minimization of SD before adjusting mean PE significantly enhances the precision of selected IOL power formulas, leading to better postoperative refractive outcomes. The effectiveness of this approach varies among different formulas, highlighting the importance of formula-specific adjustments.
Presenting Author
H. John Shammas, MD
Purpose
The purpose of this prospective study is to analyze the accuracy of the new Barrett TAL (BTAL) formula when used with the Argos measurements, and to compare the results to the commonly used Barrett Universal II formula (BU II).
Methods
This multi-center prospective study analyzes the accuracy of the new BTAL formula for IOL power calculations with the Argos sum-of-segments axial length measurements. We evaluated the percentage of eyes that achieve a postoperative manifest spherical equivalent refraction ? 0.250D, ? 0.50 D, ? 0.75D and ? 1.00D, the Mean Predictive Error and its standard deviation and the Mean Absolute Error in the entire series as well as in the short and long eyes. All values will be compared to the potential results had the BU II formula been used instead of the BTAL formula.
Results
The analysis of all eyes showed a MPE of 0.00�0.27 D with BTAL compared to 0.00�0.28 D with BU II (p=0.06). In the long eyes (AL>24.50 mm), MPE was - 0.02�0.24 with BTAL vs -0.05�0.25 D with BU II (p=0.16) and a PE within 0.50 D of 94.4% vs 93.67%. In the medium eyes (AL22.50-24.50 mm), MPE was 0.02�0.26 D with BTAL versus 0.00�0.22 with BU II (p=0.70) and a PE ?0.50 D of 92.45% vs 93.26%. In the short eyes(AL<22.50 mm),="" the="" btal="" statistically="" outperformed="" bu="" ii="" (p="0.05)" with="" a="" mpe="" of="" -0.05�0.31="" d="" vs="" 0.07�0.36="" d="" and="" a="" higher="" percentage="" of="" eyes="" achieving="" a="" refractive="" prediction="" within="" 0.50="" d="" (90.41%="" vs="">
Conclusion
The BTAL formula performed as well as BU II in the medium and long eyes and it was statistically superior (p=0.05) in the short eyes when used with sum-of-segments biometry.
Presenting Author
Greg Hazen, MD
Co-Authors
Jason Jennett BSc
Purpose
To evaluate the predictive accuracy of a new IOL calculator (J-CALC) compared to seven different IOL formulae when using keratometric (K) values from three different diagnostic devices.
Methods
Single-center, retrospective, non-interventional study of 1267 eyes that previously underwent cataract surgery and had preoperative biometry with both 2nd and 3rd generation LED topography (Ambient and Cassini, Cassini Technologies, B.V.), and optical biometry (Lenstar 900; Haag-Streit) or SS-OCT (Argos, Alcon Vision LLC). J-CALC is based on the theoretical model of the eye and only uses 3 variables to perform IOL calculation. Mean prediction error (MPE) within � 0.50D and �1.00D and standard deviations were analyzed using data from 1M post op visit.
Results
In 976 eyes (Group A) data were obtained from optical biometer (Lenstar 900); in 291 eyes (Group B) data was collected from SS-OCT (Argos). In Group A, MPE within �0.50D was as follows: J-CALC 87.6%, Barrett (modified) 73.2%, H1 72.5%, Haigis 73.4%, HofferQ 68.6%, SRK/T 68.1% (p<0.0001). in="" group="" b,="" mpe="" within="" �0.50d="" was="" as="" follows:="" j-calc="" 75.9%,="" barrett="" 71.8%,="" h1="" 67.4%,="" haigis="" 63.9%,="" hofferq="" 60.8%,="" srk/t="" 61.2%="" (p="0.016)." within="" group="" b,="" argos="" k="" values="" were="" also="" substituted="" with="" k="" values="" from="" 2nd="" and="" 3rd="" generation="" led="" topography="" (cassini).="" using="" j-calc,="" mpe="" within="" �0.50d="" was="" 70.8%="" for="" 2nd="" generation="" cassini="" and="" 74.2%="" for="" 3rd="" generation="" ambient,="" a="" significant="" improvement="" of="" +3.4%="">< />
Conclusion
J-CALC provided the highest percentage of predictive accuracy when compared to other IOL formulae, while outperforming them across all 3 keratometric device measurements. The Ambient Total K is more accurate than its predecessor and is comparable to Lenstar 900 and Argos biometers.
Presenting Author
Anuja Janakiraman, MBBS, MS
Purpose
To evaluate the predictive accuracy of the Barrett II universal(BU II)formula in patients undergoing cataract surgery with Hydrophobic acrylic foldable single piece IOLs by comparing it with Holladay Hoffer Q and SRK -T formulae and to analyze the formula's performance in eyes amongst emmetropes, myopes, and hyperopes.
Methods
Prospective study.Inclusion -202 subjects in a year aged 50-70 years,with Immature cataract with nuclear sclerosis-1,2.Exclusion-glaucoma,uveitis,retinal and corneal pathologies,intraoperative complications. Methods-IOL power was generated using the 4 formulae. All patients underwent phacoemulsification and were implanted with IOL based on Barrett II universal formula. Patients were divided into three groups based on their Axial Length : Group 1 -22�24.5 mm , Group 2:AL <22 mm, Group 3- >24.5 mm.Post operative refraction was done on days 30 , 60, and 90. IOL Power across all formulae was compared and evaluated based on the uncorrected visual acuity measured at day 90 in all the 3 groups
Results
The mean IOL power in Group 1 was 20.87 D, in Group 2 was 23.01 D,and in Group 3 was 16.31 D when BU-II formula was used . With Hoffer Q formula: 21 D,26.35 D and 14.78 D , SRK -T formula : 21.18 D,26.2 D and 15.06 D,Holladay-I formula:21.07 D,26.73 D and 14.6 D were obtained.The mean IOL power in emmetropes across all 4 formulae were similar.In hyperopes,the mean IOL power as per BU II formula was 2-3 D less compared to the other 3 formulae . In myopes,mean IOL powers per BU II was 1-2 D higher than the other 3 formulae. At day90-LogMAR UCVA:0.13 in emmetropes,0.22 in hyperopes and 0.05 in myopes. This result was compared with predicted IOL power and error using other formulae
Conclusion
119 emmetropes,40 hyperopes and 43 myopes were implanted IOL based on BU II formula and they had mean postoperative LogMar UCVA- 0.13,0.22 and 0.05 respectively .A statistically significant difference was noted between the mean IOL powers calculated using the other formula. BU II is accurate in myopes,emmetropes and better than other formulae.
Presenting Author
Jinchul KIM, MD
Co-Authors
Ronald Melles MD
Purpose
To compare the spherical equivalent refraction prediction accuracy of five modern and five traditional IOL calculation formulas when using total keratometry (TK) compared to standard keratometry (K) in normal eyes.
Methods
Retrospective study of 2554 eyes implanted with either a TFNT00 or CNWTT0 (Alcon) non-toric multifocal IOL. The standard deviation (SD) and mean absolute deviation (MAD) of the spherical equivalent prediction outcomes using the IOLMaster 700 (Zeiss) biometer K versus TK were compared using the Holladay-Wilcox heteroscedastic R statistical package for five modern IOL formulas (Barrett Universal II, EVO 2.0, Cooke K6, Kane, and PEARL-DGS) and five traditional IOL formulas (Haigis single optimization (Haigis-S), Haigis triple optimization (Haigis-T), Hoffer Q, Holladay 1, and SRK/T).
Results
A single random eye was chosen for analysis, resulting in a dataset of 1814 eyes. The SD and MAD of the spherical equivalent prediction error were statistically lower using TK compared to K, respectively, for each of the newer IOL formulas (Barrett SD 0.312 vs 0.330, EVO 2.0 SD 0.305 vs 0.312, Cooke K6 SD 0.317 vs 0.325, Kane SD 0.312 vs 0.318, PEARL-DGS SD 0.317 vs 0.324; p < 0.01 for each comparison). However, no statistically significant differences were found for the traditional IOL formulas (Haigis-SO SD 0.326 vs 0.327, Haigis-TO SD 0.325 vs 0.327, Hoffer Q SD 0.360 vs 0.356, Holladay 1 SD 0.348 vs 0.349, and SRK/T SD 0.376 vs 0.379; p > 0.05 for each comparison).
Conclusion
When used with modern IOL formulas, TK improves spherical equivalent refractive prediction in normal eyes. Therefore, TK can be reliably applied for all cataract surgery, including both normal eyes and those with previous corneal refractive surgery.
Presenting Author
Nir Sorkin, MD
Co-Authors
Adi Abulafia PhD, MD, Rotem Zadok BA, Hashem Totah BA, Filomena Ribeiro PhD, Oliver Findl MBA, MD, FEBO, Dante Buonsanti MD, Miguel Raimundo MD, Giacomo Savini MD
Purpose
To evaluate prediction accuracy of formulas included in the ESCRS-Online-IOL-Calculator using standard keratometry (K) or total keratometry (TK)
Methods
A retrospective case-series in a hospital-based academic practice, including 523 cataract patients (523 eyes). Outcome Measures: trimmed-means of the spherical equivalent prediction error (SEQ-PE, trueness), precision and absolute SEQ-PE (accuracy) of all seven formulas available on the ESCRS-Online-IOL-Calculator as well as the mean (Mean-All) and median (Median-All) of the predicted SEQ refraction of all formulas. Sub-group analyses evaluated the effect of axial length on formula accuracy.
Results
Trimmed-mean SEQ-PE of all formulas varied from -0.075 to +0.071D for K-based and from -0.003 to +0.147D for TK-based calculations, with TK-based being more hyperopic in all formulas. Precision ranged from 0.210 to 0.244D for both K-based and TK-based calculations. Absolute SEQ-PE ranged from 0.211 to 0.239D for K-based and from 0.218 to 0.255D for TK-based calculations. All formulas, including Mean-All and Median-All, showed high accuracy with 84-90% of eyes having SEQ-PEs within 0.50D. Myopic trimmed-mean SEQ-PEs significantly different from zero were observed in long eyes for Pearl DGS, Hill RBF and Hoffer QST, and in short eyes for EVO 2.0, Kane, Hoffer QST, Mean-All and Median-All.
Conclusion
Prediction accuracy of all ESCRS IOL Calculator formulas was high and globally comparable. TK-based calculations did not increase prediction accuracy and tended towards hyperopia. Observations indicating formula superiority in long and short eyes merit further evaluation.
Presenting Author
Morgan Micheletti, MD, ABO, FACS
Co-Authors
David Cooke MD, Kamran Riaz MD
Purpose
To compare the refractive predictability of two swept-source OCT-based biometry devices in eyes with low preoperative keratometric astigmatism (?1.5 D) implanted with toric intraocular lenses (IOLs) after cataract surgery.
Methods
In this retrospective study, IOL power calculations were performed using IOLMaster 700 (Group A, Carl Zeiss Meditec, Jena, Germany) and Argos (Group B, Alcon, Fort Worth, TX). Mean absolute errors (MAEs), percentage of eyes with a prediction error (PE) within �0.25 D, �0.50 D and �1.00 D, and postop cylinder (cyl) were compared between biometers. Moreover, astigmatism vector analysis and double-angle plot calculations were performed. Inclusion/exclusion criteria were applied to eliminate all possible comorbidities.
Results
A cohort of 1,032 cases were retrospectively reviewed. Postop cyl was significantly lower in group A, i.e. 0.36 � 0.36 D (centroid postop cyl: 0.04D@ 168� � 0.51D), compared to group B, i.e. 0.42 � 0.37 D (0.05D@ 59� � 0.56D) (p=0.005). Moreover, MAE was 0.313 � 0.304 D in group A and 0.322 � 0.282 in group B (p=0.251). The percentage of eyes with a PE within �0.25 D, �0.50 D and �1.00 D was 56% vs 48% (p=0.015), 80% vs 78% (p=0.592) and 96% vs 97% (p=0.122) in groups A and B, respectively.
Conclusion
Both groups achieved excellent refractive outcomes after cataract surgery. The MAE and the percentage of eyes with a PE within �0.50 D and �1.00 D were similar between both groups, however, the use of IOLMaster 700 resulted in a significantly lower postop cyl and percentage of eyes with a PE within �0.25 D.
Presenting Author
Kamran M. Riaz, MD, ABO
Co-Authors
David Cooke MD, Morgan Micheletti MD
Purpose
To compare the accuracy of two biometry techniques for cataract surgery with implantation of toric multifocal intraocular lenses (TMF-IOLs) in subjects with prior laser vision correction (LVC)
Methods
Two groups receiving TMF-IOLs using different methods for keratometry values were compared in this retrospective study. In the TK group, total keratometry using IOLMaster 700 (Carl Zeiss Meditec, Jena, Germany) with Barrett True-K formula was was used for TMF-IOL power and meridian placement calculation. In the AK group, anterior keratometry using Argos (Alcon, Fort Worth, TX) with Barrett True-K formula was used. Postop cylinder (cyl), mean absolute vector errors (MAvEs), the percentage of eyes with a vector prediction error (PE) within �0.25 D, �0.50 D and �1.00 D, astigmatism vector analysis and double-angle plot calculations were assessed in both groups.
Results
A total of 222 cases including 111 cases in TK group and 111 cases in K group were reviewed. Postop cyl was significantly lower for TK group (0.35 � 0.42 D) compared with AK group (0.46 � 0.34 D, p=0.008). Moreover, centroid postop cyl was 0.04 D @44� � 0.55D in TK group and 0.12 D @ 124� � 0.57D in AK group. The MAE was 0.35 � 0.30 D in TK group and 0.40 � 0.30 in AK group (p=0.218). The percentage of eyes with a vector PE within �0.25 D, �0.50 D and �1.00 D was 32%, 66% and 91% in TK group and 26%, 64% and 95% in AK group, with no statistical differences between both groups.
Conclusion
Both biometry techniques provided excellent refractive outcomes after cataract surgery in subjects with prior LVC. However, the total keratometry with IOLMaster 700 yielded a significantly better postop cylinder compared to anterior keratometry with Argos.
Presenting Author
William F. Wiley, MD, ABO
Co-Authors
Jason Jennett BSc, Shamik Bafna MD
Purpose
To evaluate the post-operative refractive outcomes of optical biometry, LED topography, and intraoperative aberrometry techniques in surgical planning following cataract surgery.
Methods
This study is a single-center, retrospective, non-interventional, observational study of 50 eyes who previously underwent cataract surgery and had a preoperative measurements with an optical biometer (IOL Master 500, Zeiss), LED topographer and guidance (Cassini Technologies, B.V.), in addition to live intraoperative aberrometry (IA) (ORA; Alcon). The primary endpoint was the percentage of eyes with MRSE ?0.50D. Secondary endpoints included back-calculated post-operative spherical equivalent and residual cylinder (using the preop corneal power), mean predication errors, uncorrected and corrected distance visual acuities, and mean surgical case time of surgical guidance vs IA.
Results
Data pool screening resulted in collection of 21 eyes in total with 1 month post op measurements. Following ORA recommended IOL power implantation, manifest refraction spherical equivalent (MRSE) was -0.41�0.54D (mean�SD) with 76% of eyes ?0.50D. Mean residual cylinder was 0.36D�0.42D, and BCDVA at ?20/20, ?20/30, ?20/40 were 67%, 90%, and 95% respectively. For back calculation, we replaced IA cylinder power with the preoperative planned cylinder power taken from either LED topography or optical biometry. Analysis of the residual cylinder with this method revealed the following percentages for residual cylinder ?0.50D: 86% for Ambient, 81% for IOL Master 500; compared to 86% for ORA.
Conclusion
Results from this pilot study could suggest that clinical astigmatic outcomes of IA may be comparable to LED topography under surgical guidance, with both trended towards improved outcomes, however statistical significance was not achieved. A prospective study approach with larger sample size is required to draw definitive conclusions.
Presenting Author
Hayoung Lee, MD
Co-Authors
Nahyun Park MD, Jae Yong Kim MD, PhD, Hun Lee PhD, MD, Kyu Sang Eah MD
Purpose
To compare the accuracy of three intraocular lens (IOL) power calculation formulas from the IOLMaster 700 (Carl Zeiss Meditec, Jena, Germany) (Barrett Universal II, SRK/T, Holladay 2) with seven of the latest formulas available on the ESCRS website (Barrett Universal II, Cooke K6, EVO 2.0, Hill-RBF 3.0, Hoffer QST, Kane, PEARL-DGS).
Methods
This retrospective study included 221 eyes from 221 patients. Postoperative spherical equivalent measurements at 6 months were used to calculate prediction errors by subtracting predicted preoperative refractive errors from actual postoperative refractive errors. After constant optimization, accuracy was evaluated using median absolute error (MedAE), mean absolute error (MAE), the range of absolute error (AE), and formula performance index (FPI). Lower MedAE and MAE, higher percentages of eyes within specified AE ranges, and higher FPI values indicated better accuracy. Linear regression analyzed correlations between prediction errors and preoperative biometric parameters.
Results
There were no statistically significant differences among the formulas regarding MedAE, MAE, or the percentage of eyes with absolute errors within 0.25, 0.50, 0.75, and 1.00 diopters(all p>0.05). The FPI was highest for Cooke K6 (0.414) and PEARL-DGS (0.406), indicating the highest accuracy for these formulas. Excluding these two, the prediction errors of the remaining eight formulas showed significant correlations with preoperative axial length and lens tilt. The prediction error of SRK/T correlated significantly with preoperative mean keratometry value, while BU II, SRK/T, Holladay 2, and Hoffer QST showed significant correlations with preoperative anterior chamber depth.
Conclusion
Although no significant differences were found among the formulas regarding MedAE, MAE, or error percentages within specific ranges, Cooke K6 and PEARL-DGS, which had the highest FPI, showed relatively higher accuracy. Prediction errors for each formula varied in their correlations with preoperative biometric parameters.
Presenting Author
Nahyun Park, MD
Co-Authors
Hoseok Chung MD, Jae Yong Kim PhD, MD, Hun Lee PhD, MD, Chung Min Lee MD
Purpose
To analyze the refractive prediction error after cataract surgery by comparing the accuracy of the intraocular lens (IOL) power calculation formulas in eyes with long axial lengths (>25.5mm) and short axial lengths (<22.5mm). also,="" to="" investigate="" the="" correlation="" between="" prediction="" errors="" of="" each="" formula="" and="" preoperative="" biometric="">
Methods
This retrospective study included 89 eyes with axial lengths >25.5mm and 74 eyes with axial lengths <22.5mm. after="" implantation="" of="" eyhance="" (icb00,="" johnson="" &="" johnson="" vision,="" santa="" ana,="" ca,="" usa)="" monofocal="" iol,="" spherical="" equivalents="" were="" measured="" 1="" month="" post-surgery.="" the="" accuracy="" of="" 11="" formulas,="" including="" barrett="" universal="" ii="" (buii),="" srk/t,="" haigis,="" holladay2="" from="" iolmaster="" 700,="" and="" buii,="" cooke="" k6,="" evo="" 2.0,="" hill-rbf="" 3.0,="" hoffer="" qst,="" kane,="" and="" pearl-dgs="" from="" the="" escrs="" website,="" was="" compared="" using="" median="" absolute="" error="" (medae),="" mae,="" absolute="" error="" range,="" and="" formula="" performance="" index="" (fpi).="" the="" correlation="" between="" prediction="" errors="" and="" biometric="" parameters="" was="" analyzed="" using="" linear="">
Results
In eyes with long axial lengths, Holladay2 had the highest MedAE at 0.360D compared to other formulas (0.273-0.348D). In eyes with short axial lengths, Holladay2 and Hoffer QST had the highest MedAE at 0.560D and 0.453D, respectively, while Cooke K6 had the lowest MedAE of 0.310D. Cooke K6 also had the highest FPI of 0.441. Correlations between prediction errors and biometric parameters were lower in eyes with long axial lengths. In eyes with short axial lengths, prediction errors of all formulas correlated with preoperative spherical equivalent and IOL power, and all except SRK/T and Cooke K6 also correlated with average corneal curvature.
Conclusion
No significant accuracy differences were found between IOLMaster 700-based formulas and ESCRS website formulas in long axial lengths, except for Holladay 2. In short axial lengths, Cooke K6 showed the best accuracy. All formulas were less accurate in short eyes, with prediction errors correlated with various preoperative biometric parameters.
Presenting Author
Luke Rebenitsch, MD, ABO
Co-Authors
Inder Singh MD
Purpose
To compare the refractive outcomes in a large cohort undergoing cataract surgery with toric intraocular lens (IOL) implantation assisted or not with the CALLISTO eye surgical assistant (Carl Zeiss, Meditec, Jena, Germany).
Methods
Mean absolute errors (MAEs), the percentage of eyes with a prediction error (PE) within �0.25D, �0.50D and �1.00D, postop spherical equivalent (SE) and cylinder (cyl) were compared between cases performed with (study group) or without (control group) surgical assistant. These outcomes were also compared in different subgroups, i.e. eyes with short, average or long axial length (AL) and eyes with against the rule (ATR), oblique, or regular (WTR) preop astigmatism. Inclusion/exclusion criteria were applied to eliminate all possible comorbidities.
Results
TBDPostop cyl was significantly lower in the study group (0.425 � 0.445 D) and in the control group (0.447 � 0.484 D, p<0.0001). the="" difference="" was="" also="" statistically="" different="" in="" the="" atr="" (p="0.002)" and="" oblique="" (p="0.024)" subgroups.="" moreover,="" mae="" was="" significantly="" lower="" in="" the="" study="" group="" (0.334="" �="" 0.342="" d="" vs="" 0.353="" �="" 0.398="" d,=""><0.0001), and="" the="" difference="" was="" significant="" for="" all="" subgroups="" (atr:="" p="0.005," oblique:="" p="0.004," wtr:="" p="0.002)." finally,="" percentages="" of="" eyes="" with="" a="" pe="" within="" �0.25d,="" �0.50d="" and="" �1.00d="" were="" also="" significantly="" better="" in="" the="" study="" group="" (49%,="" 79%="" and="" 97%,="" respectively)="" compared="" to="" the="" study="" group="" (47%,="" 77%,="" and="" 97%,="">
Conclusion
The use of the CALLISTO eye system resulted in significant improvements in terms postop refractive outcomes after cataract surgery with toric IOL implantation, including cyl, MAE and percentage of eyes with a PE within �0.25D, �0.50D and �1.00D.
Presenting Author
Andrea C Santiago Leon, MD
Co-Authors
Sonia Yoo MD, Robert O'Brien PhD, Ellen Koo MD
Purpose
This study evaluates the application of a novel intraocular lens (IOL) formula to predict the refractive outcome of cataract surgery patients with a history of radial keratotomy.
Methods
A retrospective chart review was done on patients undergoing IOL implantation with a history of radial keratotomy; this included implantation with phacoemulsification or IOL exchange, performed by one academic surgeon between 2017 and 2023. Multiple regression analysis was used to create the new formula using the IOL power and the number of radial keratotomy incisions in the cornea. Post-operative month one spherical equivalent (SE) was compared to the predicted SE obtained from biometry (IOL Master 500 or 700; Carl Zeiss, Germany), using the Barrett True-K, SRK/T, or Holladay 2 formulas, and to the predicted spherical equivalent using the novel formula.
Results
A total of 84 eyes were analyzed. The average subject age was 68 (�7) years. Using the novel formula, 17 patients were 0.25 diopters within the predicted target (22.0%, N=77), while 35 were within 0.5 diopters of the predicted target at one month postoperatively (45.5%, N=77). On the other hand, 10 patients were within 0.25 diopters of the IOL master predicted target (13.3%, N=75), and 15 patients were within 0.5 diopters of the IOL predicted target at one month postoperatively (20.0%, N=75).
Conclusion
Our results show a higher percentage of eyes within 0.5 diopters of the novel formula target than the IOL predicted target at post-operative month one for the IOL model and power implanted. This novel IOL formula may improve the refractive outcome accuracy of eyes undergoing cataract surgery with a history of radial keratotomy.
Presenting Author
Ali Masoudi, MD
Co-Authors
Kevin Miller MD
Purpose
To evaluate the accuracy of intraoperative aberrometry using the Alcon WaveTec Optiwave Refractive Analysis (ORA) system for IOL power refinement in eyes with a history of corneal refractive surgery, and to compare it to standard preoperative IOL power calculation.
Methods
This was a retrospective study of cataract surgeries performed at UCLA by 3 surgeons between December 2016 and December 2023. One of two methods (either the ASCRS post refractive calculator or the Barrett True K formula) was used for preoperative IOL power calculation. Prediction errors were determined for the 2 preoperative IOL power calculation methods and for the ORA-recommended IOL power, and the results were compared.
Results
A total of 3,911 procedures were performed by all UCLA surgeons. 394 procedures with history of refractive surgery that had standard selection methods from 3 surgeons, 277 (70.0%) had undergone myopic LASIK or PRK, 70 (17.7%) had undergone hyperopic LASIK or PRK, and 47 (11.9%) had undergone RK. Prediction errors �0.5 D for eyes with a history of myopic LASIK or PRK were 73.2% by pre-op calculation vs 71.4% with ORA. Prediction errors �0.5 D for eyes with a history of hyperopic LASIK or PRK were 65.7% by pre-op calculation vs 70.0% with ORA. Finally, prediction errors for eyes with a history of RK were 38.2% by pre-op calculation vs 48.9% with ORA.
Conclusion
Pre-op calculations and ORA results are comparable with the formulas used by the surgeons in this study. Post RK eyes are especially difficult to calculate regardless of the power calculation method. Detailed results comparing the ASCRS post refractive calculator and the Barrett True K formula to ORA will be reported during the live presentation.
Presenting Author
Jes�s-Enrique Arreola-Martinez, MD
Co-Authors
Roberto Gonzalez-Salinas MD, PhD, Claudia Corredor Ortega MD, Ana Laura Villavicencio-Igartua MD
Purpose
To compare the refractive predictive error employing 4 different widely used formulas in cases of rock hard cataracts using the ARGOS Biometer with Image Guidance.
Methods
This retrospective study analyzed the clinical records of 36 patients (72 eyes) with LOCS III cataracts ?NO4NC4 who underwent phacoemulsification with IOL placement. All measurements were performed using ARGOS Biometer between July and October 2023. Data included demographic details and both pre- and post-surgical outcomes, evaluating the accuracy of four widely used formulas (Holladay II, Barrett Universal II, SRKT and Kane formulas) in predicting final refractive error (FRE). Results were meticulously compared to determine which formula offered the highest accuracy.
Results
The study population had a mean age of 70.72 � 9.36 years, with a female predominance of 67%. The analysis showed that the Holladay II formula was significantly better in predicting the final refractive error (FRE) of less than 0.5 diopters compared to the subsequent best formula (64% vs 52%; p=0.023). Furthermore, it showed that no formula was significantly better in predicting the residual cylinder of less than 0.5 diopters compared to the rest (64% vs 61%, p=0.19). Axial length emerged as a crucial factor influencing the precision of these formulas, particularly in patients with hard cataracts.
Conclusion
Results differ from literature, with Holladay II outperforming the rest of the formulas in predicting final refractive error, suggesting an advantage in cases of hard cataracts (a significant observation as SS-OCT biometry is increasingly being used) The retrospective nature of the study and interoperator variability are our limitations.
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 2025 Annual Meeting registration.
