Deborah Broadbent

Sensitivity and specificity of photography and direct ophthalmoscopy in screening for sight threatening eye disease: the Liverpool diabetic eye study

Abstrac Objective: To evaluate different methods for community based screening for sight threatening diabetic eye disease. Design: Prospective study. Setting: Mobile screening unit visiting inner city community clinics; hospital assessment clinic (tertiary centre). Subjects: 395 diabetic patients registered with four general practices in an inner city location. Interventions: Community based photography with mydriasis and direct ophthalmoscopy through dilated pupils by an experienced ophthalmologist, both compared with reference standard of slit lamp biomicroscopy by a consultant specialist in medical retinal disease. Main outcome measures: Sensitivity and specificity of screening method and prevalence of sight threatening diabetic eye disease (moderate preproliferative retinopathy, circinate maculopathy, exudate within 1 disc diameter of fixation, other diabetes related eye disease). Results: 358 subjects underwent photography, 326 attended hospital clinic for ophthalmoscopy, and six were ungradable on photographs and biomicroscopy, leaving 320 for analysis. Of these 295 (91%) attended clinic within four months of photography. Sensitivity of detection of eye disease by photography was 89% (95% confidence interval 80% to 98%), significantly better than for direct ophthalmoscopy (65% (51% to 79%)). Analysis of patients with false negative results indicated possible improvement of photographic sensitivity to 93% by addition of stereoscopic macular pair photographs. Specificity of detection of sight threatening eye disease was 86% (82% to 90%) for photography and 97% (95% to 99%) for direct ophthalmoscopy. Conclusions: Since high sensitivity is essential for an effective screening programme, a photographic method should be considered as preferred option in national, community based screening programmes. Even in the hands of an experienced ophthalmologist, direct ophthalmoscopy is limited by weaknesses inherent to the instrument.

Incidence of sight-threatening retinopathy in patients with type 2 diabetes in the Liverpool Diabetic Eye Study: a cohort study.

BACKGROUND Incidence data on which to base targets and protocols for screening for sight-threatening diabetic retinopathy are few. We aimed to investigate yearly and cumulative incidence of any retinopathy, maculopathy, and sight-threatening diabetic retinopathy in patients with type 2 diabetes in an established systematic programme and to calculate optimum screening intervals according to retinopathy grade at baseline. METHODS We investigated all patients with type 2 diabetes registered with enrolled general practices (except those who were attending an ophthalmologist) who had retinopathy data available at baseline and at least one further screening event. To screen patients, we used non-stereoscopic three-field mydriatic photography and modified Wisconsin grading. Sight-threatening diabetic retinopathy was defined as moderate preproliferative retinopathy or worse, or clinically significant maculopathy in either or both eyes. FINDINGS Results were obtained from 20 570 screening events. Yearly incidence of sight-threatening diabetic retinopathy in patients without retinopathy at baseline was 0.3% (95% CI 0.1-0.5) in the first year, rising to 1.8% (1.2-2.5) in the fifth year; cumulative incidence at 5 years was 3.9% (2.8-5.0). Rates of progression to sight-threatening diabetic retinopathy in year 1 by baseline status were: background 5.0% (3.5-6.5), and mild preproliferative 15% (10.2-19.8). For a 95% probability of remaining free of sight-threatening diabetic retinopathy, mean screening intervals by baseline status were: no retinopathy 5.4 years (95% CI 4.7-6.3), background 1.0 years (0.7-1.3), and mild preproliferative 0.3 years (0.2-0.5). INTERPRETATION A 3-year screening interval could be safely adopted for patients with no retinopathy, but yearly or more frequent screening is needed for patients with higher grades of retinopathy.

Cost effectiveness analysis of screening for sight threatening diabetic eye disease

Abstract Objective: To measure the cost effectiveness of systematic photographic screening for sight threatening diabetic eye disease compared with existing practice. Design: Cost effectiveness analysis Setting: Liverpool. Subjects: A target population of 5000 diabetic patients invited for screening. Main outcome measures: Cost effectiveness (cost per true positive) of systematic and opportunistic programmes; incremental cost effectiveness of replacing opportunistic with systematic screening. Results: Baseline prevalence of sight threatening eye disease was 14.1%. The cost effectiveness of the systematic programme was £209 (sensitivity 89%, specificity 86%, compliance 80%, annual cost £104 996) and of the opportunistic programme was £289 (combined sensitivity 63%, specificity 92%, compliance 78%, annual cost £99 981). The incremental cost effectiveness of completely replacing the opportunistic programme was £32. Absolute values of cost effectiveness were highly sensitive to varying prevalence, sensitivity and specificity, compliance, and programme size. Conclusion: Replacing existing programmes with systematic screening for diabetic eye disease is justified.

Incidence of sight-threatening retinopathy in Type 1 diabetes in a systematic screening programme.

Aim To measure the cumulative incidence of any retinopathy, maculopathy and sight‐threatening diabetic retinopathy (STDR), and calculate optimal screening intervals by retinopathy grade at baseline for patients with Type 1 diabetes attending an established systematic retinal screening programme.

Prevalence of diabetic eye disease in an inner city population: The liverpool diabetic eye study

Purpose To measure the population prevalence of diabetic eye disease in an inner city setting.Methods As part of a systematic screening programme all adult diabetic patients in four general practices were invited to attend for slit-lamp biomicroscopy by a retinal specialist. Data on non-attenders were available from community-based photography.Results Of 395 diabetic patients identified, 326 attended biomicroscopy with photographic data available on a further 31, giving a 90% compliance rate. Point prevalence of diabetes in the target population was 12.4/ 1000. Demographic data included: mean age 60 years (range 13-92 years); type of control: type I 49, type II insulin-requiring (IR) 40, type II non-insulin-requiring (NIR) 268. Prevalences were as follows: any retinopathy: of all diabetic patients 33.6%, type I 36.7%, type II IR 45.0%, type II NIR 31.3%; proliferativel advanced: all 1.1%, type I 2.0%, type II IR 0, type II NIR 1.1%; clinically significant macular oedema: all 6.4%, type I 2.3%, type II IR 16.2%, type II NIR 5.7%. The percentage of patients with retinopathy requiring follow-up by an ophthalmologist was 4.5%, and 9.2% had macular exudates within 1 disc diameter of fixation or significant circinate maculopathy. Sight-threatening diabetic eye disease (STED) was found in 13.4%. A visual acuity of ≤ 6/24 in the better eye occurred in 12 (3.4%) patients and of ≤ 6/60 in the better eye in 3 (0.8%). Conclusions Compared with previous population studies, prevalences appear to have declined in type I, but remain high in type II diabetic patients and especially in those requiring insulin.

Prevalence of diabetic eye disease in patients entering a systematic primary care-based eye screening programme.

Aims Large‐scale, baseline prevalence measurements in a population at the institution of systematic retinal screening are currently unavailable. We report the prevalence of all grades of retinopathy at entry into a systematic primary care‐based diabetic eye screening programme.

Automated Segmentation of Foveal Avascular Zone in Fundus Fluorescein Angiography

PURPOSE. To describe and evaluate the performance of a computerized automated segmentation technique for use in quantification of the foveal avascular zone (FAZ). METHODS. A computerized technique for automated segmentation of the FAZ using images from fundus fluorescein angiography (FFA) was applied to 26 transit-phase images obtained from patients with various grades of diabetic retinopathy. The area containing the FAZ zone was first extracted from the original image and smoothed by a Gaussian kernel (sigma = 1.5). An initializing contour was manually placed inside the FAZ of the smoothed image and iteratively moved by the segmentation program toward the FAZ boundary. Five tests with different initializing curves were run on each of 26 images to assess reproducibility. The accuracy of the program was also validated by comparing results obtained by the program with the FAZ boundaries manually delineated by medical retina specialists. Interobserver performance was then evaluated by comparing delineations from two of the experts. RESULTS. One-way analysis of variance indicated that the disparities between different tests were not statistically significant, signifying excellent reproducibility for the computer program. There was a statistically significant linear correlation between the results obtained by automation and manual delineations by experts. CONCLUSIONS. This automated segmentation program can produce highly reproducible results that are comparable to those made by clinical experts. It has the potential to assist in the detection and management of foveal ischemia and to be integrated into automated grading systems.

Convolutional Neural Networks for Diabetic Retinopathy

Abstract The diagnosis of diabetic retinopathy (DR) through colour fundus images requires experienced clinicians to identify the presence and significance of many small features which, along with a complex grading system, makes this a difficult and time consuming task. In this paper, we propose a CNN approach to diagnosing DR from digital fundus images and accurately classifying its severity. We develop a network with CNN architecture and data augmentation which can identify the intricate features involved in the classification task such as micro-aneurysms, exudate and haemorrhages on the retina and consequently provide a diagnosis automatically and without user input. We train this network using a high-end graphics processor unit (GPU) on the publicly available Kaggle dataset and demonstrate impressive results, particularly for a high-level classification task. On the data set of 80,000 images used our proposed CNN achieves a sensitivity of 95% and an accuracy of 75% on 5,000 validation images.

Spontaneous resolution of diabetic macular oedema after discontinuation of thiazolidenediones

Aim  To report a case of spontaneous resolution of diabetic maculopathy with cystoid macular oedema in a diabetic patient after discontinuation of rosiglitazone, documented with serial ocular computed tomography (OCT) images.

Current status of screening for diabetic retinopathy in the UK

This article reviews the current status of retinopathy screening schemes in the UK. There is evidence that high‐quality diabetic retinopathy screening schemes are in existence but provision is patchy. Many health authorities have ad hoc screening programmes reaching only about 60% of patients, with unacceptable or undocumented efficacy and minimal quality control. Several models of screening are currently in use with the current preferred option being camera‐based screening. Digital imaging systems offer the best prospects for image acquisition, although at present evidence of adequate effectiveness only exists for 35 mm film‐based systems. The final report of the National Diabetic Retinopathy Screening Programme commissioned by the UK National Screening Committee for inclusion into the national service framework for diabetes, is thus eagerly awaited and should set standards for screening programmes, in order to improve the care of all those with diabetes. Quality assurance will be the main driver in the immediate future of improvements in screening programmes. Research data will provide the evidence to refine techniques and set targets in the longer term, with the emphasis on cost‐effectiveness and quality of life.