Accuracy of Toric Intraocular Lens Calculations Using Estimated Versus Measured Posterior Corneal Astigmatism.

PURPOSE: To compare the prediction accuracy of toric intraocular lens calculations using estimated vs measured posterior corneal astigmatism (PCA). DESIGN: Retrospective case series. METHODS: A total of 110 eyes of 110 patients with uncomplicated toric intraocular lens implantation were included in this study. Predicted postoperative refractive astigmatism was calculated with the Barrett Toric Calculator using the estimated PCA (E-PCA), the measured IOLMaster 700 PCA (I-PCA), and the measured Pentacam PCA (P-PCA). Refractive astigmatism prediction errors (RA-PEs), including their trimmed (tr-) centroid (mean vector), spread (precision), tr-mean absolute RA-PE (accuracy), and percentage within a certain threshold, were determined using vector analysis and compared between groups. SETTING: University Eye Clinic, Maastricht University Medical Center+, the Netherlands. RESULTS: The tr-centroid RA-PEs of the E-PCA (0.02 diopter [D] at 82.2°), the I-PCA (0.08 D at 35.5°), and the P-PCA (0.09 D at 69.1°) were significantly different from each other (P < .01), but not significantly different from zero (P = .75, P = .05, and P = .05, respectively). The E-PCA had the best precision (tr-mean 0.40 D), which was not significantly lower than the I-PCA (0.42 D, P = .53) and P-PCA (0.43 D, P = .06). The E-PCA also had the best accuracy (0.40 D), which was not significantly different from the I-PCA (0.42 D, P = .26) and significantly better than the P-PCA (0.44 D, P < .01). The precision and accuracy of the I-PCA did not significantly differ from those of the P-PCA. There were no statistically significant differences in the percentage of eyes within a certain absolute RA-PE threshold. CONCLUSIONS: The Barrett Toric Calculator using the E-PCA, I-PCA, or P-PCA showed a comparable prediction of postoperative refractive astigmatism in standard clinical practice.

Development of machine learning models to predict posterior capsule rupture based on the EUREQUO registry.

PURPOSE: To evaluate the performance of different probabilistic classifiers to predict posterior capsule rupture (PCR) prior to cataract surgery. METHODS: Three probabilistic classifiers were constructed to estimate the probability of PCR: a Bayesian network (BN), logistic regression (LR) model, and multi-layer perceptron (MLP) network. The classifiers were trained on a sample of 2 853 376 surgeries reported to the European Registry of Quality Outcomes for Cataract and Refractive Surgery (EUREQUO) between 2008 and 2018. The performance of the classifiers was evaluated based on the area under the precision-recall curve (AUPRC) and compared to existing scoring models in the literature. Furthermore, direct risk factors for PCR were identified by analysing the independence structure of the BN. RESULTS: The MLP network predicted PCR overall the best (AUPRC 13.1 ± 0.41%), followed by the BN (AUPRC 8.05 ± 0.39%) and the LR model (AUPRC 7.31 ± 0.15%). Direct risk factors for PCR include preoperative best-corrected visual acuity (BCVA), year of surgery, operation type, anaesthesia, target refraction, other ocular comorbidities, white cataract, and corneal opacities. CONCLUSIONS: Our results suggest that the MLP network performs better than existing scoring models in the literature, despite a relatively low precision at high recall. Consequently, implementing the MLP network in clinical practice can potentially decrease the PCR rate.

Accuracy of intraocular lens calculations in eyes with keratoconus.

PURPOSE: To compare the prediction accuracy of the Barrett True-K for keratoconus with standard formulas (SRK/T, Barrett Universal II, and Kane) and the Kane keratoconus formula. SETTING: Shaare Zedek Medical Center, Jerusalem, Israel, and University Eye Clinic, Maastricht, the Netherlands. DESIGN: Multicenter retrospective case series. METHODS: Eyes with stable keratoconus undergoing cataract surgery were included. The predicted refractions were calculated for SRK/T, Barrett Universal II, Barrett True-K for keratoconus (predicted and measured), Kane, and Kane adjusted for keratoconus formulas. Primary outcomes were prediction error (PE), absolute error (AE), and percentage of eyes with PE ±0.25 diopters (D), ±0.50 D, and ±1.00 D. Subgroup analyses were performed based on the severity of the keratoconus. RESULTS: 57 eyes were included in the study. The PE was not significantly different from zero for SRK/T, Barrett True-K (predicted and measured), and Kane keratoconus formulas (range 0.09 to 0.22 D, P > .05). The AE of Barrett True-K predicted (median 0.14 D) and Barrett True-K measured (median 0.10 D) were significantly lower from Barrett Universal II (median 0.47 D) and Kane (median 0.50 D), P < .001. CONCLUSIONS: The Barrett True-K formulas for keratoconus had higher prediction accuracy as compared with new generation formulas and a similar prediction accuracy as compared with the Kane keratoconus formula.

Anesthesia techniques and the risk of complications as reflected in the European Registry of Quality Outcomes for Cataract and Refractive Surgery.

PURPOSE: To determine the trends in anesthesia techniques for cataract surgery over the past decade and their relationship to surgical complications. SETTING: Clinics affiliated with the European Registry of Quality Outcomes for Cataract and Refractive Surgery (EUREQUO). DESIGN: Retrospective cross-sectional register-based study. METHODS: Variables include patient demographics, visual acuity, ocular comorbidities, surgery characteristics, intraoperative complications, and postoperative complications for the study period from January 2008, to December 2018. The anesthesia methods registered in the EUREQUO and included in the study are topical, combined topical and intracameral, sub-Tenon, regional, and general anesthesia. Multivariate logistic regression models for each complication were constructed to estimate the adjusted odds ratio (OR) and 95% CIs. RESULTS: Complete data were available of 1 354 036 cataract surgeries. Topical anesthesia increased significantly over time (from 30% to 76%, P < .001). Sub-Tenon and regional anesthesia decreased (from 27% and 38% to 16% and 6%, respectively, P < .001), and general and combined topical and intracameral anesthesia remained stable (around 2%). Sub-Tenon (OR, 0.80; 95% CI, 0.71-0.91, P < .001), regional (0.74; 95% CI, 0.71-0.78, P < .001), general (0.53; 95% CI, 0.50-0.56, P < .001), and intracameral anesthesia (0.76; 95% CI, 0.64-0.90, P = .001) carried a significantly decreased risk of posterior capsule rupture (PCR), with and without dropped nucleus, compared with topical anesthesia. The risk of endophthalmitis was significantly lower with regional anesthesia compared with topical anesthesia (OR, 0.60; 95% CI, 0.44-0.82, P = .001). CONCLUSIONS: The use of topical anesthesia for cataract surgery increased over time. Topical anesthesia is associated with an increased risk of PCR with and without dropped nucleus, and endophthalmitis.

Outcomes of cataract surgery complicated by posterior capsule rupture in the European Registry of Quality Outcomes for Cataract and Refractive Surgery.

PURPOSE: To analyze the outcomes of cataract surgery complicated by posterior capsule rupture (PCR). SETTING: European clinics affiliated to the European Registry of Quality Outcomes for Cataract and Refractive Surgery (EUREQUO). DESIGN: Retrospective cross-sectional register-based study. METHODS: Data were retrieved from the EUREQUO between January 1, 2008, and December 31, 2018. The database consists of data on demographics, intraoperative complications such as PCR, type of intraocular lens (IOL) material, postoperative refraction, corrected distance visual acuity (CDVA), and postoperative complications. RESULTS: 1 371 743 cataract extractions with complete postoperative data were reported in the EUREQUO. In 12 196 cases (0.9%), PCR was reported. After PCR, patients were more likely to receive a poly(methyl methacrylate) IOL (5.2% vs 0.4%, respectively) or no IOL (1.1% vs 0.02%, respectively) compared with patients without PCR. The refractive and visual outcomes in patients with PCR were significantly worse than in those without PCR (mean CDVA 0.13 ± 0.21 vs 0.05 ± 0.16 logMAR, P < .001; mean absolute biometry prediction error 1.15 ± 1.60 diopters [D] vs 0.41 ± 0.45 D, P < .001). A multivariate linear regression analysis, adjusting for potential explanatory variables, confirmed a statistically significant difference (0.04 logMAR, P < .001, and .70 D, P < .001, respectively). Patients with PCR had significantly more postoperative complications (corneal edema 0.88% vs 0.17%, adjusted odds ratio [aOR], 2.80 95% CI, 2.27-3.45, endophthalmitis 0.11% vs 0.02%, aOR, 4.40 95% CI, 2.48-7.81, uncontrolled intraocular pressure 0.55% vs 0.03%, aOR, 14.58 95% CI, 11.16-19.06, P < .001). CONCLUSIONS: Patients with PCR had significantly worse visual and refractive outcomes and more postoperative complications than patients without PCR. However, most of these patients achieved better postoperative visual acuity than that preoperatively.

Risk factors for posterior capsule rupture in cataract surgery as reflected in the European Registry of Quality Outcomes for Cataract and Refractive Surgery.

PURPOSE: To analyze the incidence and risk factors for posterior capsule rupture (PCR) in cataract surgery. SETTING: European clinics affiliated with the European Registry of Quality Outcomes for Cataract and Refractive Surgery (EUREQUO). DESIGN: Retrospective cross-sectional register-based study. METHODS: Data were obtained from the EUREQUO. The database contains data on demographics, comorbidities, and intraoperative complications, including PCR for the study period from January 1, 2008, to December 31, 2018. Univariate and multivariate logistic regression analyses were performed to estimate the (adjusted) odds ratio (OR) and 95% confidence intervals (CIs). RESULTS: We analyzed EUREQUO registry data of 2,853,376 patients, and 31,749 (1.1%) cataract surgeries were complicated by a PCR. Data were available of 2 853 376 patients, and 31 749 (1.1%) cataract surgeries were complicated by a PCR. The PCR rate ranged from 0.60% to 1.65% throughout the years, with a decreasing trend (P < .001). The mean age of the PCR cohort was 74.8 ± 10.5 years, and 17 29 (55.5%) patients were female. Risk factors most significantly associated with PCR were corneal opacities (OR 3.21, 95% CI, 3.02-3.41, P < .001), diabetic retinopathy (OR 2.74, 95% CI, 2.59-2.90, P < .001), poor preoperative visual acuity (OR 1.98, 95% CI, 1.88-2.07, P < .001), and white cataract (OR 1.87, 95% CI, 1.72-2.03, P < .001). CONCLUSIONS: Risk factors for PCR were identified based on the EUREQUO, and the incidence of this complication is decreasing over time.