Abstract
Objectives
The main objective of the national screening programme is to reduce the risk of sight loss among people with diabetes due to diabetic retinopathy (DR).
Methods
Offering two-field mydriatic digital photographic screening to all people with diabetes in England over the age of 12 years.
Stage of development
The programme is in its infancy, receiving the first year's annual reports from approximately 96 screening programmes, each of which have developed to offer screening to a minimum number of 12,000 people with diabetes, which would cover a population of 350,000 people with 3.4% diabetes prevalence. The national programme has commenced the External quality assurance (QA) programme in order to achieve and sustain the highest possible standards.
Potential benefits
England has a population of two million people with diabetes over the age of 12 and it is believed that there is a prevalence of blindness of 4200 and an annual incidence of blindness of 1280 people with diabetes. This programme has the potential to reduce the prevalence of blindness in England from 4200 people to 1000 people and a conservative estimate of reducing the annual incidence of DR blindness by one-third would save 427 people per annum from blindness. These figures are based on the UK certification of blindness but if World Health Organization (WHO) definitions are used the prevalence, incidence and potential reductions in blindness are much greater.
AIM of the programme
History of the programme
A reduction of diabetes-related blindness by at least one-third was declared a primary objective for Europe in 1989 (St Vincent Declaration). A number of reports1–6 subsequently supported the introduction of a national screening programme for sight-threatening DR and, in 2000, a Group commissioned by the national screening committee, proposed the introduction of a systematic national screening programme based on digital photography.7,8
The Wilson and Junger criteria for a screening programme, which are the 1968 principles applied by the WHO, have formed the basis of the national screening committee criteria for appraising the viability, effectiveness and appropriateness of a screening programme. 9 Applying these principles to sight-threatening DR provides the following evidence base. 10
The evidence that sight-threatening DR is an important public health problem.
In 1990-1991, DR was the most important cause of people registered blind among people of working age in England and Wales. 4 This finding was confirmed 11 from registrations in England and Wales in 1999-2000 and in Europe. 12
The evidence that the incidence of sight-threatening DR is going to remain the same or become an even greater public health problem.
A global epidemic of diabetes (particularly Type 2) has been reported.13–15
The evidence that sight-threatening DR has a recognizable latent or early symptomatic stage.
Studies of epidemiology,16,17 screening,18,19 and treatment of DR, 20 and treatment of diabetes 21 have shown clear evidence that sight-threatening DR has a recognizable latent or early symptomatic stage in both Type 1 and 2 diabetes.
The evidence that treatment for sight-threatening DR is effective and agreed universally?
The evidence that laser treatment is effective.
There is good evidence22–25 that photocoagulation treatment is of benefit in preventing severe visual loss, in eyes with proliferative retinopathy and that focal photocoagulation26–28 of ‘clinically significant’ diabetic macular oedema substantially reduces the risk of visual loss. Further reports29–31 have shown favourable longterm visual results of photocoagulation.
There is also evidence that DR can be prevented or the rate of deterioration reduced by improved control of blood glucose32–38 and blood pressure.39–41
A suitable and reliable screening test is available, acceptable to both healthcare professionals and (more importantly) to the public.
Studies of mydriatic digital photography against a recognized reference standard of seven-field stereophotography or an ophthalmologist using slit-lamp biomicroscopy have shown showed consistently good results.42–46 Two screening studies44–46 have demonstrated sensitivities of >87% and specificities of >86% with low ungradable image rates of <4.4%. In one of these studies, 46 the reference standard of slit-lamp biomicrocopy by an experienced ophthalmologist was validated against seven-field stereophotography. 47
The costs of screening and effective treatment of sight-threatening DR are balanced economically in relation to total expenditure on health care - including the consequences of leaving the disease untreated.
There have been reports of computer simulation models of screening for sight-threatening DR48,49 based on the health systems in the USA and Sweden, both concluding that screening for sight-threatening DR was cost-effective. An English study 50 reported results for an organized screening programme using 35 mm retinal photography and demonstrated this to be more cost-effective than the previous system of opportunistic screening. The English national screening programme used estimates 51 of costs for a theoretical population of 500,000.
Scope of the programme
To offer all people with diabetes over the age of 12 annual eye examinations for DR;
To advise on the processes to ensure that all persons with diabetes are identified securely and included in a screening programme;
To detect a high proportion of sight-threatening DR;
To ensure that there is training and accreditation available for the individual competencies needed by personnel involved in the programme, so that they can develop their skills, demonstrated through accreditation;
To treat diabetic eye disease effectively and within the nationally recommended timescales;
To ensure cost-effectiveness of the programme by best use of resources;
To set standards and to develop a quality assurance system to allow programme monitoring;
To review evidence annually in order to make recommendations for improvements in standards;
To integrate screening and treatment for DR with other aspects of diabetes care (e.g. hyperglycaemia, hypertension and hyperlipidaemia) in order to ensure optimal management of diabetes;
To involve patients individually and collectively in the development and evolution of the programme;
To develop patient information and support patient education programmes so that they can understand the strengths and weaknesses of the screening programme and the process involved;
To monitor the programme and produce annual reports on programme performance and progress;
To generate and mobilize knowledge from the data collected and from the experience of patients and clinicians in order to make continual improvements to the programme and
To identify areas where evidence is lacking and to facilitate research in these areas.
Coverage
The latest Quality and Outcomes Framework data 52 for 2006-2007 (QOF) have shown a national prevalence of diabetes in England of 3.7% of the population (1,961,976 people with diabetes aged ≥17 years).
Based on an analysis of costs, skills and demand to produce a practical size for planning a screening programme for DR, the national recommendation, and one of the quality standards, is that a screening programme should include a minimum number of 12,000 people with diabetes, which would therefore cover a minimum population of 350,000 people (based on 3.4% with diabetes). This has resulted in the development of approximately 96 screening programmes in England to offer annual digital photographic screening to all people with diabetes over the age of 12 years.
One critical factor for all screening programmes is to make sure that they have a regularly updated and accurate list of patients to invite for screening and that there is careful management of patients that do not require an invitation (e.g. patients who are terminally ill or are under the care of an Ophthalmologist and regularly having their retinae examined).
There is an urgent need for electronic updates between General Practice and Screening Programmes of patients with diabetes who need to be invited for screening as manual methods are prone to errors. In some parts of the country large differences have been seen in the number of patients with diabetes reported as part of the QOF data and the number known to the screening programme. There is clearly a significant risk of an adverse event in a screening programme if patients are not being invited for screening because the programme is unaware of their existence.
Quality assurance of the programme
Quality Assurance in the DR Screening Programme is being lead by a National QA Director, who published a pilot study 53 of 10 existing services, to inform the development of the recommended Quality Assurance Objectives and Standards, supported by professional Regional QA Managers and, linking closely with the Strategic Health Authorities who have the responsibility for the quality of screening programmes.
Internal quality assurance is carried out by the programme itself with monitoring of statistical output data, which is included in the programme's annual report that is submitted annually to the national team.
An External on-line QA set of images is being developed so that all graders working within the screening programme will grade a certain number of these images per annum and the results compared with consensus grading results.
Peer review visits are being organized at intervals to assure the quality of the DR screening programme.
The expected reduction in blindness due to diabetic retinopathy
Registration as blind or partially sighted in England and Wales is voluntary and is initiated by certification by a Consultant Ophthalmologist and the forms, which are now termed Certificates of Visual Impairment (full or partial), are sent to the Office of Population Censuses and Surveys. The definition of blindness and partial sight in the UK is shown in the box below:
The WHO definition of blindness is the same as that of ‘Legal blindness’ in the United States and Europe, i.e. a best corrected visual acuity in the better eye ≤6/60.
In 1984-1994, the reported rates 54 of 10-year incidences of blindness were 1.8%, 4.0% and 4.8% in persons in the younger-onset, older-onset taking insulin and older-onset not taking insulin groups, respectively. Improvements in diabetic control have resulted in improvements 55 and in some areas marked improvements56–59 in incidence of blindness, particularly where organized screening programmes have been introduced. The best documented reduction of blindness has come from Iceland 60 where the prevalence of legal blindness (WHO definition VA <6/60 in the better seeing eye) from DR in the diabetic population dropped from 2.4% in 1980 to 0.5% in 2005 (to approximately one-fifth of the original). Iceland, as an example, has shown that screening compliance 61 and optimal management is very effective in preventing and/or reducing severe complication of diabetes.
A reasonable estimate of the prevalence and incidence of blindness due to DR in England is derived from a study 62 in Fife in Scotland where they found the prevalence of blindness due to diabetes was 0.21% of the diabetic population and the incidence of blindness due to diabetes was 0.064% (SD: 0.024, 95% CI: 0.049-0.079) diabetic population/year. These figures are lower than European figures because of the voluntary nature of the certification and the more stringent criteria for registration.
Extrapolating these rates to the two million people with diabetes in England gives a prevalence of blindness of 4800 and an annual incidence of blindness of 1280 people. If one was to reduce the prevalence of blindness due to diabetes to one-fifth of the original, as has been achieved in Iceland, this has the potential to reduce the prevalence of blindness in England from 4800 people to under 1000 people. Even if the annual incidence is only reduced by one-third, as recommended by the St Vincent Declaration, this would save 427 people per annum from blindness.
In order to overcome some of the problems of inadequate data collection in England, the English national screening programme is collecting visual acuity data on all patients with diabetes that are known to the screening programme. Data collection facilities are good within screening programmes but not in Ophthalmology departments where data collection relating to visual acuity or any other aspect of care of a patient with DR is very sparse at present and mechanisms need to be put in place to provide outcome data for this population-based screening programme, so that the anticipated success of the national screening programme and of treatment facilities within Ophthalmology can be measured.