Moorfields Eye Hospital NHS Foundation Research Articles

Feasibility of Automated Deep Learning Design for Medical Image Classification by Healthcare Professionals with Limited Coding Experience

Background: Deep learning has huge potential to transform healthcare however significant expertise is required to train such models. In this study, we therefore sought to evaluate the use of automated deep learning software to develop medical image diagnostic classifiers by healthcare professionals with limited coding - and no deep learning - expertise. Methods: We used five publicly available open-source datasets: (i) retinal fundus images (MESSIDOR); (ii) optical coherence tomography (OCT) images (Guangzhou Medical University/Shiley Eye Institute, Version 3); (iii) images of skin lesions (Human against Machine (HAM)10000) and (iv) both paediatric and adult chest X-ray (CXR) images (Guangzhou Medical University/Shiley Eye Institute, Version 3 and the National Institute of Health (NIH)14 dataset respectively) to separately feed into a neural architecture search framework that automatically developed a deep learning architecture to classify common diseases. Sensitivity (recall), specificity and positive predictive value (precision) were used to evaluate the diagnostic properties of the models. The discriminative performance was assessed using the area under the precision recall curve (AUPRC). In the case of the deep learning model developed on a subset of the HAM10000 dataset, we performed external validation using the Edinburgh Dermofit Library dataset. Findings: Diagnostic properties and discriminative performance from internal validations were high in the binary classification tasks (range: sensitivity of 73·3-97·0%, specificity of 67-100% and AUPRC of 0·87-1). In the multiple classification tasks, the diagnostic properties ranged from 38-100% for sensitivity and 67-100% for specificity. The discriminative performance in terms of AUPRC ranged from 0·57 to 1 in the five automated deep learning models. In an external validation using the Edinburgh Dermofit Library dataset, the automated deep learning model showed an AUPRC of 0·47, with a sensitivity of 49% and a positive predictive value of 52%. The quality of the open-access datasets used in this study (including the lack of information about patient flow and demographics) and the absence of measurement for precision, such as confidence intervals, constituted the major limitation of this study. Interpretation: All models, except for the automated deep learning model trained on the multi-label classification task of the NIH CXR14 dataset, showed comparable discriminative performance and diagnostic properties to state-of-the-art performing deep learning algorithms. The performance in the external validation study was low. The availability of automated deep learning may become a cornerstone for the democratization of sophisticated algorithmic modelling in healthcare as it allows the derivation of classification models without requiring a deep understanding of the mathematical, statistical and programming principles. Future studies should compare several application programming interfaces on thoroughly curated datasets. Funding Statement: National Institute for Health Research, United Kingdom. PAK is supported by an NIHR Clinician Scientist Award (NIHR-CS--2014-14-023). The research was also supported by the National Institute for Health Research (NIHR) Biomedical Research Centre based at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology. Declaration of Interests: The authors have no conflict of interest to declare. JL and TB are employees of DeepMind Technologies, a subsidiary of Alphabet Inc. RC is an intern at DeepMind. PK is an external consultant for DeepMind. Ethics Approval Statement: Not required.

Comparative analysis of 2 swept-source optical coherence tomography biometers

To report the level of agreement, repeatability, and correlation of axial length (AL), mean keratometry, central corneal thickness, anterior chamber depth, lens thickness, and corneal diameter measurements of 2 swept-source optical coherence tomography (SS-OCT) biometers, the IOLMaster 700 (reference biometer) and the Argos (new biometer). Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom. Retrospective case series. Each patient had SS-OCT biometry with the reference biometer and new biometer. In addition to reporting the statistical results derived from both eyes, this study included a subgroup analysis of right eyes and left eyes. The level of agreement between the biometers was represented with the Bland-Altman method. Power vector analysis of the J0 (Jackson cross-cylinder, axes at 0 degrees and 90 degrees) and J45 (Jackson cross-cylinder, axes at 45 degrees and 135 degrees) vectorial components of astigmatism was performed. Internal consistency was assessed with the Cronbach α coefficient of reliability. The dispersion of probability distribution was computed with the coefficient of variation. The intraoperator repeatability was calculated using the intraclass correlation coefficient (ICC). The study comprised 112 patients (218 eyes). There was a statistically significant difference between the 2 biometers in all measurements (P<.05) except AL. The level of correlation between the reference biometer and the new biometer was very high for all the parameters except corneal diameter, and the agreement was high. The ICC and internal consistency were excellent with both biometers. The new biometer provided good agreement and repeatability compared with the reference biometer.

The incidence and clinical outcome of complications in 4,000 consecutive strabismus operations

To test the validity of the British Ophthalmic Surveillance Unit (BOSU) study's incidence figure of severe complications following strabismus surgery and to determine the incidence, type, risk factors, and outcome of all strabismus surgery complications at a single institution. A prospective audit of consecutive strabismus operations performed by consultants or trainees was carried out between 2011 and 2016 at Moorfields Eye Hospital NHS Foundation Trust. Patient diagnosis, age, sex, surgical details, complications, and outcome were recorded from hospital records. We classified complications as minor, moderate, or severe. The outcome was graded using the Bradbury and Taylor grading system (I to IV), with a poor or very poor outcome meaning loss of corrected visual acuity or unexpected primary position diplopia. A total of 4,076 consecutive strabismus operations were performed during the study period. There were 46 (1.13%) complications, of which 28 (0.69%) were minor, 7 (0.17%) were moderate, and 9 (0.22%) were severe. Only 1 patient (0.02%) had a poor visual outcome. Two patients had nonocular postoperative complications (0.05%). In this large, prospective series, we found the rate of severe complications of strabismus surgery to be 1 in 455 cases. Our results validate the findings of the BOSU study.

Intravitreal aflibercept compared with panretinal photocoagulation for proliferative diabetic retinopathy: the CLARITY non-inferiority RCT

Background Panretinal photocoagulation (PRP) has been the standard of care for patients with proliferative diabetic retinopathy (PDR) for the last 40 years. It prevents severe visual loss in PDR but is also associated with adverse effects on visual functions. Objectives The clinical efficacy and mechanistic evaluation of aflibercept for proliferative diabetic retinopathy (CLARITY) trial evaluated the clinical efficacy, mechanisms and cost-effectiveness of intravitreal aflibercept (Eylea®, Regeneron, Tarrytown, NY, USA/Bayer Pharma AG, Berlin, Germany therapy for PDR. Design A multicentre, prospective, individually randomised, single-masked, active-controlled trial with concurrent economic evaluation that tested the non-inferiority of intravitreal aflibercept versus standard care PRP at 52 weeks. A subset of the participants enrolled in a mechanistic evaluation substudy. Setting 22 UK NHS clinical sites. Participants Patients aged at least 18 years having either treatment-naive PDR or active retinal neovascularisation (NV) despite prior PRP requiring treatment and best corrected visual acuity (BCVA) of 54 Early Treatment Diabetic Retinopathy Study (ETDRS) letters or better in the study eye were included. Eyes with evidence of macular oedema at baseline confirmed by central subfield thickness &gt; 320 µm on spectral-domain optical coherence tomography were excluded. Intervention In the intervention arm, intravitreal aflibercept injections were given at baseline, 4 and 8 weeks and patients were subsequently reviewed every month and injected pro re nata based on the treatment response defined by degree of regression of retinal NV. In the comparator arm, PRP was completed in 2-weekly sessions and then supplemented if necessary at 8-weekly intervals. Main outcome measures The primary outcome was the mean change in BCVA at 52 weeks utilising a linear mixed-effects model incorporating data from both week 12 and week 52. Results A total of 232 participants (116 per arm) were recruited between August 2014 and November 2015. A total of 221 and 210 participants contributed to the intention-to-treat (ITT) model and per-protocol (PP) analysis, respectively. Economic evaluation was undertaken on 202 participants (101 per arm) with complete cost and outcome data. Aflibercept was non-inferior and superior to PRP in both the ITT population [mean BCVA difference 3.9 letters, 95% confidence interval (CI) 2.3 to 5.6 letters; p &lt; 0.0001] and the PP population (difference 4.0 letters, 95% CI 2.4 to 5.7 letters; p &lt; 0.0001). From a public sector multiagency perspective that covers health and social care services, treatment with aflibercept costs more in terms of total resource use (mean adjusted total additional cost per patient = £5475, bootstrapped 95% CI £5211 to £5750) than PRP over a 12-month follow-up period. There were a small number of important safety events in each arm. Patients were more satisfied with aflibercept than PRP. Limitations This study is limited to 1 year of follow-up. Conclusions At an additional cost, the study shows that intravitreal aflibercept is an effective alternative treatment option for PDR in the first year. Future work Future research is needed to evaluate the long-term benefits of aflibercept in comparison with PRP and other anti-vascular endothelial growth factor agents for this condition. Trial registration Current Controlled Trials ISRCTN32207582. Funding This project was funded by the National Institute for Health Research (NIHR) Efficacy and Mechanistic Evaluation programme, a Medical Research Council and NIHR partnership. Aflibercept was supplied by Bayer Plc (Reading, UK). The study was sponsored by NIHR Moorfields Biomedical Research Centre and supported by the UK Clinical Research Network. The research was supported by the NIHR Biomedical Research Centre at Moorfields Eye Hospital NHS Foundation Trust and University College London Institute of Ophthalmology, the NIHR Moorfields Clinical Research Facility and the UK Clinical Reasearch Collaboration-registered King’s Clinical Trials Unit at King’s Health Partners, which is partly funded by the NIHR Biomedical Research Centre for Mental Health at South London and Maudsley NHS Foundation Trust and King’s College London.

Combining optical coherence tomography with visual field data to rapidly detect disease progression in glaucoma: a diagnostic accuracy study.

Progressive optic nerve damage in glaucoma results in vision loss, quantifiable with visual field (VF) testing. VF measurements are, however, highly variable, making identification of worsening vision ('progression') challenging. Glaucomatous optic nerve damage can also be measured with imaging techniques such as optical coherence tomography (OCT). To compare statistical methods that combine VF and OCT data with VF-only methods to establish whether or not these allow (1) more rapid identification of glaucoma progression and (2) shorter or smaller clinical trials. Method 'hit rate' (related to sensitivity) was evaluated in subsets of the United Kingdom Glaucoma Treatment Study (UKGTS) and specificity was evaluated in 72 stable glaucoma patients who had 11 VF and OCT tests within 3 months (the RAPID data set). The reference progression detection method was based on Guided Progression Analysis™ (GPA) Software (Carl Zeiss Meditec Inc., Dublin, CA, USA). Index methods were based on previously described approaches [Analysis with Non-Stationary Weibull Error Regression and Spatial enhancement (ANSWERS), Permutation analyses Of Pointwise Linear Regression (PoPLR) and structure-guided ANSWERS (sANSWERS)] or newly developed methods based on Permutation Test (PERM), multivariate hierarchical models with multiple imputation for censored values (MaHMIC) and multivariate generalised estimating equations with multiple imputation for censored values (MaGIC). Ten university and general ophthalmology units (UKGTS) and a single university ophthalmology unit (RAPID). UKGTS participants were newly diagnosed glaucoma patients randomised to intraocular pressure-lowering drops or placebo. RAPID participants had glaucomatous VF loss, were on treatment and were clinically stable. 24-2 VF tests with the Humphrey Field Analyzer and optic nerve imaging with time-domain (TD) Stratus OCT™ (Carl Zeiss Meditec Inc., Dublin, CA, USA). Criterion hit rate and specificity, time to progression, future VF prediction error, proportion progressing in UKGTS treatment groups, hazard ratios (HRs) and study sample size. Criterion specificity was 95% for all tests; the hit rate was 22.2% for GPA, 41.6% for PoPLR, 53.8% for ANSWERS and 61.3% for sANSWERS (all comparisons p ≤ 0.042). Mean survival time (weeks) was 93.6 for GPA, 82.5 for PoPLR, 72.0 for ANSWERS and 69.1 for sANSWERS. The median prediction errors (decibels) when the initial trend was used to predict the final VF were 3.8 (5th to 95th percentile 1.7 to 7.6) for PoPLR, 3.0 (5th to 95th percentile 1.5 to 5.7) for ANSWERS and 2.3 (5th to 95th percentile 1.3 to 4.5) for sANSWERS. HRs were 0.57 [95% confidence interval (CI) 0.34 to 0.90; p = 0.016] for GPA, 0.59 (95% CI 0.42 to 0.83; p = 0.002) for PoPLR, 0.76 (95% CI 0.56 to 1.02; p = 0.065) for ANSWERS and 0.70 (95% CI 0.53 to 0.93; p = 0.012) for sANSWERS. Sample size estimates were not reduced using methods including OCT data. PERM hit rates were between 8.3% and 17.4%. Treatment effects were non-significant in MaHMIC and MaGIC analyses; statistical significance was altered little by incorporating imaging. TD OCT is less precise than current imaging technology; current OCT technology would likely perform better. The size of the RAPID data set limited the precision of criterion specificity estimates. The sANSWERS method combining VF and OCT data had a higher hit rate and identified progression more quickly than the reference and other VF-only methods, and produced more accurate estimates of the progression rate, but did not increase treatment effect statistical significance. Similar studies with current OCT technology need to be undertaken and the statistical methods need refinement. Current Controlled Trials ISRCTN96423140. This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 22, No. 4. See the NIHR Journals Library website for further project information. Data analysed in the study were from the UKGTS. Funding for the UKGTS was provided through an unrestricted investigator-initiated research grant from Pfizer Inc. (New York, NY, USA), with supplementary funding from the NIHR Biomedical Research Centre at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology, London, UK. Imaging equipment loans were made by Heidelberg Engineering, Carl Zeiss Meditec and Optovue (Fremont, CA, USA). Pfizer, Heidelberg Engineering, Carl Zeiss Meditec and Optovue had no input into the design, conduct, analysis or reporting of any of the UKGTS findings or this work. The sponsor for both the UKGTS and RAPID data collection was Moorfields Eye Hospital NHS Foundation Trust. David F Garway-Heath, Tuan-Anh Ho and Haogang Zhu are partly funded by the NIHR Biomedical Research Centre based at Moorfields Eye Hospital and UCL Institute of Ophthalmology. David F Garway-Heath's chair at University College London (UCL) is supported by funding from the International Glaucoma Association.

Methods for assessing forward and backward light scatter in patients with cataract

To compare objective methods for assessing backward and forward light scatter and psychophysical tests in patients with cataracts. Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom. Prospective case series. This study included patients scheduled for cataract surgery.Lens opacities were grouped into predominantly nuclear sclerotic, cortical, posterior subcapsular, and mixed cataracts. Backward light scatter was assessed using a rotating Scheimpflug imaging technique (Pentacam HR), forward light scatter using a straylight meter(C-Quant), and straylight using the double-pass method (Optical Quality Analysis System, point-spread function [PSF] meter). The results were correlated with visual acuity under photopic conditions as well as photopic and mesopic contrast sensitivity. The study comprised 56 eyes of 56 patients. The mean age of the 23 men and 33 women was 71years (range 48 to 84years). Two patients were excluded. Of the remaining, 15patientshad predominantly nuclear sclerotic cataracts, 13 had cortical cataracts, 11 had posterior subcapsular cataracts, and 15 had mixed cataracts. Correlations between devices were low. The highest correlation was between PSF meter measurements and Scheimpflug measurements (r=0.32). The best correlation between corrected distance visual acuity was with the PSF meter (r=0.45). Forward and backward light-scatter measurements cannot be used interchangeably. Scatter as an aspect of quality of vision was independent of acuity. Measuring forward light scatter with the straylight meter can be a useful additional tool in preoperative decision-making.

Automated analysis of retinal imaging using machine learningtechniques for computer vision.

There are almost two million people in the United Kingdom living with sight loss, including around 360,000 people who are registered as blind or partially sighted. Sight threatening diseases, such as diabetic retinopathy and age related macular degeneration have contributed to the 40% increase in outpatient attendances in the last decade but are amenable to early detection and monitoring. With early and appropriate intervention, blindness may be prevented in many cases. Ophthalmic imaging provides a way to diagnose and objectively assess the progression of a number of pathologies including neovascular ("wet") age-related macular degeneration (wet AMD) and diabetic retinopathy. Two methods of imaging are commonly used: digital photographs of the fundus (the 'back' of the eye) and Optical Coherence Tomography (OCT, a modality that uses light waves in a similar way to how ultrasound uses sound waves). Changes in population demographics and expectations and the changing pattern of chronic diseases creates a rising demand for such imaging. Meanwhile, interrogation of such images is time consuming, costly, and prone to human error. The application of novel analysis methods may provide a solution to these challenges. This research will focus on applying novel machine learning algorithms to automatic analysis of both digital fundus photographs and OCT in Moorfields Eye Hospital NHS Foundation Trust patients. Through analysis of the images used in ophthalmology, along with relevant clinical and demographic information, DeepMind Health will investigate the feasibility of automated grading of digital fundus photographs and OCT and provide novel quantitative measures for specific disease features and for monitoring the therapeutic success.

Is strabismus the only problem? Psychological issues surrounding strabismus surgery

Strabismus, which has been estimated to occur in 4% of the adult population, can have a negative psychosocial effect on an individual's quality of life. While correction of strabismus may improve quality of life, this is not always the case, even if realignment surgery is clinically successful. Surgeons need to understand patients' expectations of postsurgical outcomes and the impact surgery has on their psychosocial well-being. This was a prospective cross-sectional study of 220 adult patients undergoing strabismus surgery at Moorfields Eye Hospital NHS Foundation Trust. Clinical assessments were performed before surgery and again 2weeks and 3months after surgery. Standardized psychological questionnaires were completed preoperatively and 3 and 6months postoperatively. Of 220 patients included in the study (mean age, 45years), 54% were female, 81% were white, and 58.6% had a concomitant deviation. Prior to surgery, 24% of study participants were experiencing clinical levels of anxiety; 11% were suffering clinical levels of depression. Quality of life improved and anxiety and depression were reduced after surgery. Approximately 6% of the sample regretted having undergone surgery either at 3 or 6months after treatment. It is important to identify patients who are experiencing significant psychosocial distress and to assess their postsurgical expectations in order to improve the outcomes of strabismus surgery. Presurgical psychosocial support should be considered for these patients.

Automated analysis of retinal imaging using machine learning techniques for computer vision

There are almost two million people in the United Kingdom living with sight loss, including around 360,000 people who are registered as blind or partially sighted. Sight threatening diseases, such as diabetic retinopathy and age related macular degeneration have contributed to the 40% increase in outpatient attendances in the last decade but are amenable to early detection and monitoring. With early and appropriate intervention, blindness may be prevented in many cases. Ophthalmic imaging provides a way to diagnose and objectively assess the progression of a number of pathologies including neovascular (“wet”) age-related macular degeneration (wet AMD) and diabetic retinopathy. Two methods of imaging are commonly used: digital photographs of the fundus (the ‘back’ of the eye) and Optical Coherence Tomography (OCT, a modality that uses light waves in a similar way to how ultrasound uses sound waves). Changes in population demographics and expectations and the changing pattern of chronic diseases creates a rising demand for such imaging. Meanwhile, interrogation of such images is time consuming, costly, and prone to human error. The application of novel analysis methods may provide a solution to these challenges. This research will focus on applying novel machine learning algorithms to automatic analysis of both digital fundus photographs and OCT in Moorfields Eye Hospital NHS Foundation Trust patients. Through analysis of the images used in ophthalmology, along with relevant clinical and demographic information, DeepMind Health will investigate the feasibility of automated grading of digital fundus photographs and OCT and provide novel quantitative measures for specific disease features and for monitoring the therapeutic success.

Automated analysis of retinal imaging using machine learningtechniques for computer vision.

There are almost two million people in the United Kingdom living with sight loss, including around 360,000 people who are registered as blind or partially sighted. Sight threatening diseases, such as diabetic retinopathy and age related macular degeneration have contributed to the 40% increase in outpatient attendances in the last decade but are amenable to early detection and monitoring. With early and appropriate intervention, blindness may be prevented in many cases. Ophthalmic imaging provides a way to diagnose and objectively assess the progression of a number of pathologies including neovascular ("wet") age-related macular degeneration (wet AMD) and diabetic retinopathy. Two methods of imaging are commonly used: digital photographs of the fundus (the 'back' of the eye) and Optical Coherence Tomography (OCT, a modality that uses light waves in a similar way to how ultrasound uses sound waves). Changes in population demographics and expectations and the changing pattern of chronic diseases creates a rising demand for such imaging. Meanwhile, interrogation of such images is time consuming, costly, and prone to human error. The application of novel analysis methods may provide a solution to these challenges. This research will focus on applying novel machine learning algorithms to automatic analysis of both digital fundus photographs and OCT in Moorfields Eye Hospital NHS Foundation Trust patients. Through analysis of the images used in ophthalmology, along with relevant clinical and demographic information, Google DeepMind Health will investigate the feasibility of automated grading of digital fundus photographs and OCT and provide novel quantitative measures for specific disease features and for monitoring the therapeutic success.

Comparative analysis of optical biometers

To evaluate the level of agreement, repeatability, and correlation of 2 optical biometers, the IOLMaster (based on partial coherence interferometry [PCI]) and Aladdin (based on optical low-coherence interferometry [OLCI]) in terms of axial length (AL), mean keratometry (K), anterior chamber depth (ACD), and corneal diameter. Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom. Prospective comparative case series. Each participant had biometry with both biometers. The level of agreement between thebiometers was calculated using the Bland-Altman method. Double-angle polar plots were used to display the astigmatism vectors. Internal consistency was computed with the Cronbach α coefficient of reliability, whereas the dispersion of probability distribution was assessed with the coefficient of variation (CoV). The intraoperator repeatability was analyzed using the intraclass correlation coefficient (ICC); the AL was evaluated in all eyes and the mean K and the ACD in a subset of eyes. The study comprised 215 eyes; the ICC for the mean K and ACD was calculated in a subset of 54 eyes. There was a statistically significant difference between the 2 biometers in all measurements (P<.05). The level of agreement was high with both methods. The ICC and internal consistency were excellent with both biometers; the CoVs were low. The OLCI biometer provided good agreement and repeatability compared with the PCI biometer, the current gold standard in ocular biometry. Further comparative studies are necessary to clearly define the role of the OLCI biometer in predicting postoperative refractive outcomes. Dr. Findl is a scientific advisor to Carl Zeiss Meditec AG. None of the other authors has a financial or proprietary interest in any material or method mentioned.

Nonpenetrating femtosecond laser intrastromal astigmatic keratotomy in eyes having cataract surgery

To describe the effect of femtosecond laser intrastromal astigmatic keratotomy (AK) performed during cataract surgery. Moorfields Eye Hospital, London, United Kingdom. Prospective case series. This study comprised patients having laser cataract surgery with concurrent astigmatism management by intrastromal AK. All eyes had greater than 0.7 corneal diopter (D) cylinder. An intrastromal AK nomogram with 8.0 mm diameter paired symmetric limbal centered arcs was used. Corneal keratometry was measured preoperatively and 1 month postoperatively using a KR8100PA topographer-autorefractor. Astigmatic analyses were performed using the Alpins method considering 3 vectors-target induced astigmatism (TIA), surgically induced astigmatism (SIA) and difference vector (DV)-and calculation of coupling measures. In all, 196 eyes of 133 patients were analyzed. The mean TIA (equivalent to preoperative corneal cylinder) was 1.21 D ± 0.42 (SD) (range 0.75 to 2.64 D) and the mean SIA was 0.74 DC ± 0.40 (range 0.00 to 2.86). The mean difference vector was 0.74 ± 0.38 D (range 0.00 to 2.25 D). The mean correction index was 0.63 ± 0.32 (range 0.00 to 1.93), indicating that the mean astigmatism correction was 63%. Fourteen eyes (7.1%) and 7 eyes (3.6%) had an astigmatism correction of greater than 100% and greater than 120%, respectively. Overall 0%, 48.5%, and 51.5% of eyes had 0.50 D or less, 1.0 D or less, or greater than 1.0 D, respectively, preoperatively compared with 32.1%, 85.7%, and 14.3%, postoperatively. There were no cases of corneal endothelial perforation or inadvertent placement within the visual axis. The intrastromal AKs were easily programmed as an integral part of laser-assisted cataract surgery without additional cost, significantly reduced corneal cylinder, and appeared to be safe through 1 month of follow-up. Dr. Day was supported by the National Institute for Health Research (NIHR) Biomedical Research Centre based at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology. Dr. Stevens has been a consultant to Optimedica Inc., now part of Abbott Medical Optics.

Risks and outcomes associated with primary intraocular lens implantation in children under 2 years old with congenital and infantile cataract: the UK and Ireland IoLunder2 cohort study

BackgroundPublic health initiatives to prevent childhood blindness have emphasised the importance of effective intervention for congenital cataract. Primary intraocular lens implantation, an important recent management innovation, has been rapidly adopted despite the absence of robust supporting evidence. The IoLunder2 study aimed to provide much needed insights into visual and adverse outcomes after surgery with and without primary intraocular lens implantation in children aged under 2 years. MethodsActive surveillance case identification of cataract surgery and systematic data collection were undertaken through the British Congenital Cataract Interest Group (comprising 173 ophthalmologists who manage affected children in the UK and Ireland), which identified more children undergoing surgery for cataract than were reported to the National Health Service statutory databases. Multivariable regression analysis was used to estimate associations of intraocular lens use with visual outcome (assessed with standard hospital-specific but harmonised protocols) and incidence of adverse events at 1 year after surgery, with adjustment for all relevant confounding factors (eg, ocular comorbidity or age at surgery), which were agreed a priori on the basis of previous research or biological plausibility. All reported odds ratios (ORs) are fully adjusted. FindingsOf 221 children, 56 of 131 with bilateral cataract and 48 of 90 with unilateral cataract had primary intraocular lens implantation. Implantation was independently associated with better visual outcome in bilateral (OR 5·4, 95% CI 1·09–26·4, p=0·04) but not unilateral disease, and with increased odds of reoperation (bilateral OR 5·5, 95% CI 2·3–13·2, p=0·002; unilateral 16·7, 4·1–68·9, p=0·009). Intraocular lens use did not reduce the odds of postoperative glaucoma, the key sight-threatening complication, after surgery in bilateral or unilateral cataract. InterpretationUse of intraocular lenses in young children, particularly in settings where follow-up is limited, should be critically reassessed. The absence of visual benefit and the lack of a previously postulated protective effect against postoperative glaucoma question the value of intraocular lens implantation in unilateral disease. The association between intraocular lens implantation and better early visual outcomes in bilateral disease needs to be balanced against the risk of reoperation and exposure to general anaesthetics during a key neurodevelopmental period. FundingUlverscroft Vision Research Group, Great Ormond Street Hospital/UCL Institute of Child Health National Institute for Health Research Biomedical Research Centre, and Moorfields Eye Hospital NHS Foundation Trust/UCL Institute of Ophthalmology National Institute for Health Research Biomedical Research Centre.

Visual health inequalities: findings from UK Biobank

BackgroundLittle is known about inequalities in eye health. Between 2009 and 2010, 117 908 UK Biobank participants (aged 40–69 years) undertook an ophthalmic assessment, which included distance visual acuity. UK Biobank was not designed to be a fully representative population sample so prevalence estimation is precluded. However, the size and diversity of the sample provide a unique opportunity for investigation of socioeconomic influences on visual health in UK adults. MethodsHabitual (usual optical correction) distance acuity was measured with a standardised computer-based system. 112 314 participants were reliably assigned, on the basis of acuity in the better eye, into one of six categories spanning the spectrum of vision from bilateral normal vision (log of minimum angle of resolution [logMAR] 0·2 or better) to low-vision–blind (≤0·5, WHO taxonomy). Socioeconomic information included educational qualifications and Townsend Index. Multinomial and ordinal regression analyses were undertaken. FindingsThe frequency of normal bilateral vision decreased with age (age 40–49 years, 86% [21 934/25 645]; 50–59 years, 77% [27 482/35 786]; and 60–70 years, 72% [36 461/50 883]). Overall, risk of visual impairment across severity categories was associated with an increasing gradient of key demographic and socioeconomic variables, indicating deprivation. These patterns of visual health inequalities were not explained by risk of underlying eye disease. For example, compared with normal vision, socially significant visual impairment (SSVI), a mid-range category of visual impairment, was associated with increasing age (risk ratio 1·05, 95% CI 1·046–1·06), being female (1·09, 1·01–1·16), no educational qualifications (1·7, 1·4–1·9), a higher deprivation score (1·08, 1·07–1·09), and being part of any minority ethnic group (eg, Asian 2·5, 2·1–2·9). Participants unable to work or unemployed were at least 30% more likely to be in the SSVI category than were those with normal vision and, if employed, at least 9% more likely to have a lower status job. InterpretationThere are consistent patterns of associations between visual impairment across the full spectrum including, importantly, people with mild impairment, and known health determinants as well as key social outcomes. To our knowledge, our study provides evidence for the first time that policies tackling health inequalities as well as initiatives to address inequalities in ophthalmological clinical settings have the potential to improve visual health outcomes. FundingThis work was funded by the National Eye Research Centre. PMC is funded by the Ulverscroft Foundation and JSR receives part funding from the National Institute for Health Research (NIHR) Biomedical Research Centre at Moorfields Eye Hospital NHS Foundation Trust and University College London Institute of Ophthalmology. The study was undertaken at University College London Institute of Child Health, which receives a proportion of its funding from the Department of Health's NIHR Biomedical Research Centres funding scheme.

Multifocal toric intraocular lenses versus multifocal intraocular lenses combined with peripheral corneal relaxing incisions to correct moderate astigmatism

To compare the outcomes of multifocal toric intraocular lens (IOL) implantation and standard nontoric multifocal IOL implantation combined with peripheral corneal relaxing incisions (PCRIs) in patients with moderate preexisting corneal astigmatism. Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom. Prospective randomized controlled trial. Patients with visually significant cataract in both eyes who desired spectacle independence and had corneal astigmatism (1.00 to 2.50 diopters [D]) had multifocal toric IOL implantation in 1 eye and a nontoric multifocal IOL combined with PCRIs (Donnenfeld nomogram) in the contralateral eye. Outcome measures were visual acuity, astigmatic vector reduction, digital toric IOL axis determination, spectacle need, and patient satisfaction. Thirty patients were enrolled. Three months postoperatively, the mean residual refractive astigmatism was 0.45 D ± 0.49 (SD) in the toric IOL group and 0.72 ± 0.61 D in the PCRI group. The mean uncorrected distance visual acuity was 0.10 ± 0.14 in the toric IOL group and 0.15 ± 0.14 in the PCRI group and the mean uncorrected near visual acuity, 0.43 ± 0.11 and 0.39 ± 0.10, respectively. The mean absolute misalignment from the intended axis was 2.52 ± 1.97 degrees. The rotation was within ±3 degrees in 90.9% of all cases and within ±6 degrees in all cases. On the questionnaire, 52.9% of all patients said they required spectacles only when reading for a long time. Although refractive astigmatism decreased in both groups, multifocal toric IOL implantation was predictable with good rotational stability. No author has a financial or proprietary interest in any material or method mentioned.

Correction of moderate corneal astigmatism during cataract surgery: Toric intraocular lens versus peripheral corneal relaxing incisions

To compare the astigmatism-reducing effect of a toric intraocular lens (IOL) and peripheral corneal relaxing incisions (PCRIs). Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom. Prospective masked bilateral randomized study. Cataract patients with a preoperative corneal astigmatism of 1.0 to 2.5 diopters (D) were included. All patients received a toric IOL in 1 eye and a nontoric IOL plus a PCRI in the other eye. Postoperative follow-up was at 1 hour, 1 month, and 6 months. The uncorrected distance visual acuity, corrected distance visual acuity, autorefraction (Topcon RM-8800), and subjective refraction were recorded. The IOL axis was assessed using retroillumination photographs. The study enrolled 60 eyes of 30 patients. The mean astigmatism vector reduction was 1.74 D ± 0.64 (SD) in the toric IOL group and 1.27 ± 0.76 D in the PCRI group; the difference was statistically significant (P=.042). The mean absolute rotation of the toric IOL was 2.5 ± 1.8 degrees (maximum 6.3 degrees) in the first 6 postoperative months. Astigmatism increased in the PCRI group between the 1-month and 6-month follow-up (mean 0.38 ± 0.27 D; maximum 1.00 D) (P<.001). Toric IOLs and PCRIs both reduced astigmatism; however, toric IOLs reduced astigmatism to a higher extent and they were more predictable.

A Review of the Medical Treatment of Pediatric Glaucomas at Moorfields Eye Hospital

To report on the medical treatments used for pediatric glaucomas. A retrospective case series consisting of reviewing the medical notes of pediatric glaucoma patients under the care of the Glaucoma Service at Moorfields Eye Hospital NHS Foundation Trust. The medical notes of 200 patients were selected. The following outcomes were assessed: (1) the use of individual medical therapies; (2) the intraocular pressure (IOP) reduction effect of individual medicines; (3) the reporting of side effects. Medicines were prescribed 1592 times (200 patients). The median % IOP reduction for latanoprost 50 mcg/mL was -17.2% and for the topical beta blockers was -17.7% (as monotherapy), with no statistical difference in IOP-lowering effect between all the medicines (P=0.19). Side effects were reported in 19.5% of all patients-the highest occurrence with brimonidine tartrate 0.2% (in 17% patients) and the lowest occurrence with the prostaglandin analogue and prostamide medicines (in 3.8% patients). The combination of dorzolamide hydrochloride 2%, timolol maleate 0.5% had the greatest persistence of 1 year. The IOP-lowering effects of all the glaucoma medicines were not significantly different although the combination of dorzolamide hydrochloride 2%, timolol maleate 0.5% had the greatest persistence.

Fluctuations in corneal curvature limit predictability of intraocular lens power calculations

To analyze fluctuations in corneal curvature over time. Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom. Case series. A 3-piece IOL was implanted in 1 eye and a 1-piece IOL in the other eye through a 3.2 mm clear corneal temporal incision. Keratometry was performed preoperatively and at several points in time postoperatively. Differences between measurements were analyzed by power vectors. Statistical significance was assessed by monovariate, bivariate, and trivariate paired t tests. Acute angle shifts were determined as differences between meridians at 2 points in time. Fifty patients were enrolled. From preoperatively to 1 year postoperatively, the changes in vector components (M, J0, J45) were, respectively, -0.02 diopter (D) ± 0.23 (SD) (P=.38), -0.07 ± 0.27 D (P=.02), and +0.04 ± 0.25 D (P=.14). Corresponding changes from 1 year to 2 years postoperatively were +0.01 ± 0.25 D (P=.73), +0.01 ± 0.23 D (P=.83), and +0.01 ± 0.16 D (P=.40). The meridian shift was -5 ± 32 degrees (P=.13) from preoperatively to postoperatively and +3 ± 22 degrees (P=.23) from 1 year to 2 years. Surgically induced astigmatism was composed of slight flattening in the horizontal meridian and slight steepening in the oblique meridian but was insignificant in relation to random fluctuations, which were almost equally large between postoperative measurements 1 year apart. The fluctuations were not due to imprecision in measurement. Dr. Norrby is a retiree from Abbott Medical Optics, Inc., and holder of a small amount of stock. Dr. Findl is scientific advisor to Abbott Medical Optics, Inc. No other author has a financial or proprietary interest in any material or method mentioned.

Effect of Fluorescein Dye Staining of the Tear Film on Scheimpflug Measurements of Central Corneal Thickness

To assess whether fluorescein dye in the tear film results in a clinically relevant change in central corneal thickness (CCT), as measured with Scheimpflug imaging (Pentacam HR; Oculus, Wetzlar, Germany). The study was conducted at Moorfields Eye Hospital NHS Foundation Trust. In this prospective, randomized, controlled, single-center study on healthy volunteers, CCT was measured using rotating Scheimpflug imaging and partial coherence interferometry technique. After 3 baseline measurements of both eyes, a drop of 0.25% fluorescein with 0.5% proxymetacaine hydrochloride dye, as used for measuring intraocular pressure with applanation Goldmann tonometry, was instilled in the first eye and measurements were repeated after 1 minute and 5, 10, 20, 40, and 60 minutes. A fluorescein drop was then instilled in the second eye, and 1 measurement was performed after 1 minute. Immediately afterwards, the second eye was rinsed with 2 ml of balanced saline solution and the CCT measurement was repeated. On a different day, a Schirmer II test was performed. Thirty eyes of 15 volunteers were included in this study. The average CCT measured with Scheimpflug imaging was 539.1 μm (SD, 32.2 μm), and after instilling fluorescein, CCT values increased by 46.6 μm (SD, 11.4 μm; P < 0.01, paired t test). In 10 of 15 subjects, the fluorescein layer was still present 40 minutes after instilling the drop. Rinsing the eye decreased the fluorescein layer by 49.6% and a value of 24.8 μm (SD, 16.8 μm). Because CCT measurements are critical for corneal refractive surgery, Scheimpflug imaging should not be performed after use of fluorescein dye for staining of the corneal epithelium or for applanation tonometry.

Posterior capsule opacification and capsular bag performance of a microincision intraocular lens

To assess the axial stability and posterior capsule opacification (PCO)-inhibiting efficacy of a microincision hydrophilic intraocular lens (IOL) (Idea 613 XC) and a standard hydrophobic open-loop acrylic IOL (Acrysof SA60AT). Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom. Clinical trial. Patients with bilateral cataract received a microincision cataract surgery (MICS) IOL in 1 eye and a small-incision cataract surgery (SICS) IOL in the contralateral eye. Follow-up examinations were performed immediately after surgery and postoperatively at 1 week and 3 and 24 months. Posterior capsule opacification was assessed using retroillumination photography. Anterior chamber depth (ACD) was measured using partial coherence laser interferometry, and IOL tilt and decentration were measured using a Purkinje meter. Seventy eyes (35 patients) were enrolled. The mean ACD at 3 months was 4.91 mm ± 0.26 (SD) in the MICS group and 4.60 ± 0.23 mm in the SICS group. The anterior IOL shift over the first 3 months was slightly smaller with the MICS IOL (81 μm) than with the SICS IOL (254 μm). The mean ACD was 0.377 ± 178 μm deeper with the MICS IOL (P<.01, paired t test). Two years postoperatively, the amount of regeneratory PCO was higher with the MICS IOL than with the SICS IOL (image analysis software score 2.6 versus 1.9; P=.005). Two years after surgery, the PCO rate was slightly higher with the MICS IOL. The 2 IOLs had similar performance in terms of axial stability, decentration, and tilt. No author has a financial or proprietary interest in any material or method mentioned. Additional disclosures are found in the footnotes.