Diabetes is a chronic condition and there are major systemic impacts if it is poorly controlled. High levels of blood glucose resulting from poor control result in damage to the vascular system causing both macrovascular and microvascular complications as summarised in Table 1. Systemic implications were outlined in more detail in Part 1 of this series.
Background
For people living with diabetes, sight loss is one of the most worrying consequences of this disease. Until recently it was the number one cause of blindness in the working age population (16-64 years).1 For the first time in 50 years diabetic eye disease has been overtaken by inherited retinal disorders as the leading cause of certifiable blindness in this age group in England and Wales. This is probably due to a number of factors including better glycaemic control as a result of the introduction of the National Service Framework for Diabetes and also the introduction of a nationwide diabetic eye screening programme.
Diabetic eye screening programmes have been set up in England, Northern Ireland, Scotland and Wales. In England the NHS diabetic eye screening programme is commissioned by NHS England; in Wales it is the Diabetic Retinopathy Screening Service for Wales (DRSSW); in Northern Ireland it is the Diabetic Retinopathy Screening Programme (DRSP); and in Scotland it is the Scottish Diabetic Retinopathy Screening (DRS) collaborative.
In 2003 the implementation of diabetic eye screening (DES) in England was announced as part of the Delivery Strategy for the National Service Framework for Diabetes and by 2008 the whole of England was covered by locally commissioned diabetic eye screening programmes. This article will look at the structure and pathways of the English Diabetic Eye Screening Programme (details of differences with the other programmes will be outlined in a later article in the series)
The need for a screening programme
Screening is used in medicine on a population who are at increased risk of a medical condition in which there may be no obvious signs or symptoms of manifest disease and the patient is unaware of having the condition. The intention is to identify early signs or markers of disease in individuals that enable that individual to receive advice, information, further testing if required and then to receive appropriate treatment and management to reduce morbidity. This reduces the risk of clinically significant disease within the population being considered.
Is diabetes eye screening an effective approach?
The main reason for screening for diabetic eye disease is to reduce the incidence and prevalence of sight loss caused by diabetes through the early detection of the retinal changes associated with this, and appropriate management to reduce the progression of the disease.
The Diabetes Control and Complications Trial (DCCT) produced evidence that improving glucose control in type 1 diabetes had a significant impact on both the development and progression of diabetes eye disease. Similarly the UK Prospective Diabetes Study (UKPDS) showed an improvement in outcomes with better control in type 2 diabetes. This last study also revealed the impact of good blood pressure control on the rate of progression of retinopathy.
One of the main criteria for establishing a disease screening service is that the disease should be treatable, with better outcomes if it can be detected at an early stage. As these studies show, improved management of diabetes and blood pressure are the first line of attack for managing diabetes eye disease and can have a significant impact of the progression of the eye disease.
In 1968 The World Health Organization published guidelines on the ‘principles of screening for disease’ as developed by Wilson and Jungner.2 The principles can be summarised under four categories. (Cochrane and Holland also suggested criteria which are included below 3). Diabetic Eye Disease more than adequately meets all of these recommendations.
Another important reason for implementing routine screening within three months of being notified of a diagnosis of diabetes is that, in the early stages of diabetic retinopathy, the condition is asymptomatic. Diabetic eye disease can progress to severe visual loss without any symptoms for the patient until the condition is well advanced, at which point treatment is far less effective and significantly more intense.
Diabetes eye screening is not designed to detect other conditions such as glaucoma, malignancy, inflammation and cataract although these conditions may be opportunistically discovered in the process of screening and can be referred on to ophthalmology for appropriate management according to local protocols. Patients should always be advised to also have the free NHS eye examination with their own optometrist. Current guidance from the College of Optometrists is that this should only be annual if the patient is not being seen annually within the screening programme or if there are other ocular conditions requiring more frequent examination.
Limitations of screening
The limitations of screening centre upon the degree to which the test can accurately distinguish between those individuals with the condition from those without it. It is not possible to detect everyone with the disease, nor exclude everyone who does not have the disease because of various factors, including observer interpretation of results, camera resolution and the effect of media changes in the eye.
The Exeter Standards are used as the basis for the accepted approach in systematic diabetic eye screening. The standards were set in 1984 at a Diabetes UK meeting and were then accepted into the National Service Framework as the methodology standard for National Screening for Diabetic Retinopathy in the UK. They recommend that any screening test for diabetic retinopathy should achieve a minimum sensitivity of 80 per cent and a minimum specificity of 95 per cent. Sensitivity is the capacity of the test to give a positive finding when the individual has the condition being sought. Specificity is the capacity of the test to give a negative finding when the individual does not have the condition being sought. Tests are compared to a reference standard, in this case either slit-lamp biomicroscopy by a medical retina specialist or 7 field stereo 35 degree photography to determine true positive, true negative, false positive and false negatives. Digital imaging meets the Exeter Standards.
Diabetes eye screening can be very stressful for patients because of the potential for sight threatening eye disease being discovered. As no screening test is perfect there will inevitably be false positives and false negatives. The different outcomes can impact upon a patient’s emotional well-being. Diabetes Eye Screening does have one notable difference from other health screening programmes. People in the target population have already been diagnosed with diabetes – a long-term condition with major implications on their lifestyle and the prognosis for their health. They are aware to some degree that management of their diabetes has an impact on the risks for the development and progression of diabetic retinopathy
Diabetes eye screening is most effective if patients attend annually on a regular basis so that early progressive changes can be detected and tracked through time. For this reason there is strict guidance on the number of patients who are seen for eye screening. 100 per cent of the eligible population must be invited to eye screening annually and a minimum of 70 per cent of those invited should attend. An achievable standard is considered to be 80 per cent. A recent initiative in the London North East programme targeting the persistent non-attenders showed that these are the most at-risk group as non-attendance at eye screening often correlates with poor management of their diabetes. It is important that patients have a positive experience during any encounter with the screening team either on the phone about their appointment, or at the screening centre itself, as this is a significant factor promoting regular attendance for screening.
Diabetic eye screening care pathway
In 2014 Public Health England refined the care pathways within diabetes eye screening. The key elements are:
- Database: Anyone with a diagnosis of diabetes mellitus aged 12 and over should be seen for annual diabetic retinopathy screening in the NHS Diabetic Eye Screening Programme (NDESP), except for people who have no perception of light (NPL) in both eyes (some other exclusions also apply). Each diabetes eye screening programme is required to hold a collated list of people with diabetes registered with a GP practice in the area covered by the programme which they must update regularly as people move into and out of area, and are newly diagnosed or are deceased.
- Appointments: Everyone on the collated list is invited annually for screening. This can either be to a fixed appointment or an invitation to arrange an appointment at a time and place of their convenience. This will vary depending on the type of screening service being offered. In inner city areas with dense populations and good transport links, fixed centrally located clinics are an effective model whereas in large rural areas with a scattered population and less frequent transport links a multi-site and/or mobile screening service can offer better access and availability of appointments. If people fail to attend the first invitation they are re-invited at least once. Evidence from the London North East DESP shows offering multiple appointment opportunities increases the uptake of screening. If someone with diabetes still does not attend their GP is notified and they are then invited again at their next annual due date.
- Visual acuities: When the patient arrives for their appointment it is important to confirm the demographic details (among many things to avoid screening the wrong patient and to avoid sending a report to the wrong address) and obtain consent for the screening procedure. They are initially pre-screened and visual acuities are recorded. Some programmes also record a brief history and symptoms including any history of attendance in the Hospital Eye Service (HES). A few programmes based in areas where the prevalence of severe hypertension is higher than average also measure blood pressure. The patient is then dilated with mydriatic eye drops.
- Digital imaging: After a suitable period of time to allow for dilation (20 minutes or more) digital images are taken. In general each eye will have a macular centred and a disc centred image taken (Figures 1 and 2). Anterior eye images are also useful if there are media changes to help identify possible causes of reduced visual acuity (because this is a factor in determining the maculopathy grade). The recent changes to the grading criteria allow for a ‘jigsaw’ approach so if the digital images are poor, taking a series of images that cover the area of retina normally seen in the macular/disc centred model is considered acceptable. The images are then stored on a server (with suitable back-ups in place) for grading at a later stage. In some cases the image quality will not be acceptable or it may not be possible to photograph the fundus (poor mobility is often a cause of this) and I will cover that later in the pathway.
[CaptionComponent="2950"]
[CaptionComponent="2951"]
- Grading: Once the digital images have been taken and the encounter closed, the patient enters the grading pathway. All images are initially reviewed by a primary grader. The recent changes to the care pathways established a ‘features-based’ grading scheme. The grader identifies retinal features and ticks the relevant boxes in the grading template which generates a nationally recognised diabetic retinopathy grade (R0, R1, R2, R3S, R3A) and maculopathy grade(M0, M1).
– R0M0: In 90 per cent of the outcomes where the primary grader detects none of the features of diabetic retinopathy the results are sent to the patient and GP and the patient will be re-invited for routine annual screening. The other 10 per cent will go forward in the grading pathway to have a secondary grade. This is part of the programme quality assurance process.
– R1, R2, M1 and non DR lesions: All cases of diabetic eye disease (excluding proliferative eye disease) detected at the primary grade will go forward in the grading pathway to a secondary grade. The secondary grader is not able to see the primary grade.
– R1M0: If the primary and secondary graders agree on this final grade the results are sent to the patient and GP and the patient will be re-invited for routine annual screening.
– Referable retinopathy: R1M1, R2M0, R2M1: If the primary and secondary graders agree on this grade the patient goes forward in the grading queue to the Referral Outcome Grader for a final assessment.
– R3S, R3A: If proliferative eye disease is detected at primary grade the patient is fast-tracked directly to the Referral Outcome Grader
– Unassessable: If the primary or secondary grader considers that the images are not of sufficient quality to be safely graded they go forward to the Referral Outcome Grader.
– Urgent Non-DR lesion: If the primary or secondary grader considers that there is a potentially sight-threatening non-diabetes related lesion they go forward to the Referral Outcome Grader
- Arbitration Grade: If the primary and secondary graders disagree on the grade it will go forward to the arbitration grader. The arbitration grader does have access to both the primary and secondary graders outcomes and is the final arbiter for the grade. The arbitration grader is in a good position to build an impression of individual graders performance.
- Referral Outcome Grade: This is considered to be the final outcome grade and this is carried out by a senior level grader supervised by the clinical lead. This grader is able to choose the final action outcome which includes referring patients to the HES, slit-lamp biomicroscopy (for unassessable images), digital surveillance (for borderline routine referable grades)or routine annual screening (for non-referable grades).
- Digital surveillance: diabetic eye screening is an annual process. If the patient needs more frequent monitoring they are seen in a parallel pathway within the programme called Digital Surveillance. Patients seen within this pathway include pregnant women who are known to be already diabetic (recalled every three months), early/borderline pre-proliferative retinopathy and early maculopathy (with good/stable visual acuities and minimal exudation).
- Referable retinopathy: Patients with referable diabetic retinopathy and referable non DR retinopathy are referred into the HES for management. Patients graded R3A are referred urgently. Patients graded with R2 and/or M1 are referred routinely. If the referral was for a non-diabetes related condition, the patient should still be recalled for annual diabetic eye screening. If the referral was for diabetic retinopathy they are not invited back for eye screening until they are discharged back to the service from the HES
Quality assurance
Quality assurance is built into the daily running of the eye screening programme. There are 19 Quality Assurance Objectives which require regular monitoring and that have to be met by a programme. Each programme’s performance is monitored against both a minimum and also an achievable percentage for these objectives.
Quality assurance includes the double grading of all images with detectable eye disease and further grading depending on outcomes. Clinical managers can then monitor grader performance and any issues highlighted can be addressed.
All screening and grading staff must be appropriately qualified in accordance with national standards to ensure that they meet the competencies in the national occupational standards. This currently means achieving the relevant sections of the City and Guilds Qualification.3
Optometrists are exempt from Units 3,4 and 5 (Table 3) through the ‘Accreditation of Prior Experiential Learning’ along with confirmation of their current GOC registration. In addition, optometrists who are undertaking the MSc modules at the City University and have achieved the unit in Diabetic Eye Disease are exempt from Unit 2 (Diabetes and its relevance to retinopathy screening)
To register for the City & Guilds qualification4 visit www.drsdiploma.org
All graders also must take part in a monthly online test and training scheme (a minimum of 10 out of 12 tests in a rolling 12-month period must be carried out). Clinical managers are able to review the results and must feedback to graders regularly. Its primary function is as a learning experience that can help identify training needs for individual graders.
The eye screening pathway includes strict timelines against which it is assessed:
- Results letters must be issued to both the patient and the GP within three weeks of the screening encounter
- Patients with active proliferative eye disease require referral within two weeks of the initial screening encounter and should be seen within the HES within two weeks of being referred
- Patients requiring routine referral should be seen within the HES within 13 weeks of the final outcome grade
- Patients who have been referred into the HES must be tracked as part of a fail-safe process to ensure that they have been seen by ophthalmology after referral and to make sure that patients do not fail between the cracks in the pathway between eye screening and ophthalmology. Quarterly feedback to the Programme Board is considered an achievable standard
- Patients with unassessable images must be seen for a slit-lamp biomicroscopy assessment within 14 weeks of the final grade.
Role of optometry within diabetes eye screening
There are various opportunities for optometrists within the National Eye Screening Programme. Some programmes, such as London North East employ optometrists as image graders, slit-lamp providers and clinical managers. In other areas diabetes eye screening is located within optometrist practices either as a stand-alone scheme or as one strand within a more multi-element programme. The level of involvement will vary depending on the way the scheme has developed over time, and also how it has been commissioned.
The College of Optometrists has issued guidance for optometrists as outlined in Table 4.
Conclusion
Diabetes Eye Screening is a quality assured national screening programme that is proving to be a very effective population based approach to tackling the sight-threatening implications of diabetes eye disease. It is complimentary to, and not an alternative to regular optometrist eye examinations. It is important that eye screening programmes make this clear to their patients. Similarly it is important that optometrists also encourage their patients to attend for annual eye screening in their local programme. In the next article in this series I will take a closer look at the grading of retinopathy.
Read more
A practitioner’s perspective: Diabetic retinopathy screening
Diabetes part 1 – disease overview
Diabetes part 3 – The English Grading Scheme
Diabetic retinopathy VRICS – part 2
References
1 Ophthalmology - A comparison of the causes of blindness certifications in England and Wales in working age adults (16–64 years), 1999–2000 with 2009–2010 Gerald Liew, Michel Michaelides, Catey Bunce.
2 Wilson JMG, Jungner G. Principles and practice of screening for disease. WHO Chronicle, 1968;22(11):473.
3 Cochrane AL, Holland WW. Validation of screening procedures. British Medical Bulletin, 1971; 27; 3–8
4 City & Guilds Level 3 Qualifications in Diabetic Retinopathy Screening (7360)
Peter Mitchell is an optometrist with the Diabetic Eye Screening Service and is clinical manager of the recently enlarged service in north-east London