Retinopathy screening is optometry involved?

Following publication of the implementation phase of the National Service Framework for diabetes, the National Screening Committee has established a Diabetic Retinopathy Screening Project Advisory Group to consider how a systematic national retinopathy screening programme will be introduced in England.
The Diabetic Retinopathy Screening Project Advisory Group was set up in 2002 with a number of terms of reference, including:

To advise the UK National Screening Committee on the implementation, development, review and modification of a national screening programme in England
To support local delivery to achieve the objective in line with the Diabetes National Service Framework delivery strategy that by March 2006, 'primary care trusts will have ensured that a minimum of 80 per cent of people with diabetes are offered screening for the early detection (and treatment if needed) of diabetic retinopathy as part of a systematic programme that meets the national standards, rising to 100 per cent coverage of those at risk of retinopathy by end 2007'.

Membership of the group covers a wide spectrum of professionals who are involved in diabetic retinopathy screening. The main work of the group has been delegated to three sub groups which have considered specific areas:

Education and training
Camera specification and IT
Grading, referral and quality assurance.

Optometry is represented on the main group by Richard Broughton and David Cartwright, with Trevor Warburton, Susan Blakeney and Geoff Roberson taking an active part in the sub groups.
There are a number of complex issues within the discussions that will have an influence on optometric participation. This article sets out the case that has been presented on behalf of optometry, the background to the ongoing discussions and finally the anticipated next steps that will affect optometry.
At the Project Advisory Group's most recent meeting in September, the optometric representatives gave a detailed presentation setting out clearly all the benefits that optometrist involvement will contribute to a national screening programme for diabetic retinopathy and how optometry can meet the proposed service objectives for quality assurance set out in the National Screening Committee's consultation document.
The principal points made in the submission to the advisory group were as follows:

In many parts of the UK, successful and effective co-management schemes have been established involving the participation of optometrists. Indeed, the majority of existing diabetic retinopathy screening services across the country involve optometrists
Clinical audits of a number of well-organised schemes have shown that optometrists trained to appropriate standards can safely and reliably refer sight-threatening eye disease to acceptable levels of sensitivity and specificity.

As an example, an audit of the East Suffolk Diabetic Eye Screening Scheme for the year April 2000 to March 2001 showed the following results:

84 per cent of the target population were screened during the year (3,268 individuals)
91 per cent of GP practices participated
96 per cent of patients had no, or minimal, background retinopathy
4 per cent (130 individuals) were referred to the hospital, of whom 49 were discharged, 76 were followed up at the hospital and 19 had laser treatment
150 patients were noted as having other ophthalmic conditions, including cataract and glaucoma (4.6 per cent).

Hulme et al in their evaluation of a district-wide screening programme for diabetic retinopathy utilising trained optometrists using slit-lamp and Volk lenses1 assessed sensitivity and specificity for an optometric scheme:

Sensitivity for STED was 87 per cent
Specificity was 91 per cent.

A guidance document on diabetic co-management has been produced by the College of Optometrists, the Royal College of Ophthalmologists, the Royal College of General Practitioners, the Royal College of Physicians and Diabetes UK (formerly the British Diabetic Association), with the support of the Department of Health. Thus it can be seen that the involvement of optometrists in the care of diabetics is well supported by all the relevant professional and special interest groups.
Feedback indicates that where successful optometric schemes are in place, there is a desire to continue those schemes, simply building in the use of digital cameras. Examples of such schemes include a business-specific model from a group of PCTs in Greater Manchester and the well-established Dorset optometric scheme.

BENEFITS OF OPTOMETRIC SCHEMES
A National Screening Committee consultation paper has proposed standards for developing quality assurance and lists 17 service objectives. Optometrist-based schemes will ensure that many of these are met in an appropriate manner and will best meet the needs of patients.
The following gives the significant service standards relating to optometric schemes and the main points that highlight the benefits of optometric involvement.

To maximise the number of invited persons accepting the test (Service objective 4). Optometrists' practices are conveniently located and easily accessible, and can offer flexible appointment times. Good accessibility will increase the percentage uptake of patients within the schemes. Together with the consequent reduction in travel time and costs for patients, this will encourage better compliance. Optometrists' practices are situated in almost all communities and there are sufficient optometrists available to provide screening services for all patients with diabetes within their local vicinity.
To ensure grading is accurate (6): There is concern to avoid tiredness leading to errors. Major causes of errors are monotony and boredom due to a repetitive workload. The mixture of optometrists' workload, which encompasses time allocated to involvement in formal screening schemes and routine optometric practice, adds interest rather than the reverse, and avoids both these elements.
To ensure timely reporting of abnormal screening results (8): Optometrists have an excellent knowledge of normal retinal appearances through their everyday work in examining large numbers of patients. Accredited optometrists therefore have the skills to grade as well as screen. Thus there is no delay between screening and grading, and the timely referral of appropriate cases is ensured. The proportion of ungradeable images could be as high as 10 per cent, depending on the definition of an ungradeable image. Almost all of the ungradeable images are from patients over 70 years of age2 and this age group should not be subject to avoidable extra appointments. The NSC will need to define what constitutes an ungradeable image for the national programme.
To ensure both GP and patient are informed of negative results (11). Optometrists are able to screen and grade at the same appointment and can therefore give the patient the results immediately, whether negative or otherwise, thus avoiding unnecessary anxiety on the part of the patient.
To minimise the anxiety associated with screening (15). See comment for (11) above.
To ensure timely rescreening (16). Optometric practices have short waiting lists and are used to operating their own recall systems, and could thus easily fit into a central recall system. Optometry can also be very flexible as to capacity, with the ability to cope with the gradually rising demand for screening over the coming years.

COMPARATIVE COSTS
Optometrist-based schemes can offer an extremely cost-effective resource to help meet the proposed service objectives as set out in the National Screening Committee's consultation paper.
It has been suggested that optometry schemes are more expensive than centralised camera schemes. However, in making this claim it does not seem to be the case that truly comparative costs are used and it is often the case that the budgets quoted for alternative schemes do not include realistic infrastructure overheads.
There is much speculation that mobile or centralised screening will be very cheap, but little evidence to support this. In 2000, in Rochdale, 4,800 patients were screened by digital camera at a cost of £126,000 in staff (£26.25 per event), plus capital costs and no account of accommodation costs.
A figure around this sum may well be appropriate for an optometric camera-based scheme, but would be likely to include overhead costs as well as screening and primary grading. The optometry fee could also be agreed as a 'per screening fee', so that failure to attend does not add to PCT costs. Some simplistic calculations, which appear to show a low cost of centralised or mobile screening, are actually a 'cost per person with diabetes' which make no allowance for 'DNAs' ('did not attend').
The costs to the patient in terms of time and travel should also not be overlooked. Many will find it more convenient to be screening locally, and the NSC itself said that 'Épatients normally managed in primary care should not necessarily be expected to travel long distances to the diabetes centre just to be screened'.
In comparing costs it is tempting to ignore the cost of buildings and overheads when an existing location is chosen. Health economists will say that this is incorrect and like must always be compared with like, so the overheads of an optometric practice must be compared with the overheads of a hospital or primary care resource centre.

AUDIT
The National Screening Committee regards quality assurance as a key issue in ensuring that the screening programme delivers a high quality service. There is a particular concern that problems encountered in screening for breast and cervical cancer should not be repeated in retinopathy screening. An important issue is the minimum numbers of patients graders will need to see to ensure statistically sound quality assurance (QA). The evidence base for deciding on appropriate minimum numbers is very limited and the initial proposals requiring graders to see a minimum of 1,000 patients a year do not appear to be based on specific evidence and if adopted would effectively exclude optometry from providing a screening service which would include primary grading.
While it is accepted that there is a requirement for quality assurance to be statistically robust, optometry has argued that it is also important that numbers are realistically achievable and do not exclude from participation well-trained professionals with the appropriate skills. Numbers should be set at realistic and pragmatic levels.
This topic has generated more discussion and debate than any other in the project advisory group. It has now been recognised that there need to be specific standards for optometric schemes. The current proposal is that an optometrist involved in screening patients within their own practice, as part of a scheme, will photograph and primary grade Ð that is determine the presence or absence of retinopathy and the presence or absence of referable retinopathy. They should see a minimum of 250 patients and if they see less than 500, they should have additional patient screening episodes, sent electronically, to grade.
For example, if the optometrist sees 250 patients they will have another 250 sets of patient images to grade. However, it is not yet agreed how many additional electronic gradings an optometrist who examines between 250 and 500 patients will have to see. The quality assurance then will be based on reviewing a proportion of decisions made on a total of 2,000 images graded (500 patient episodes, times four pictures per patient).
A significant minority of patients who attend for screening by digital photography will produce an upgradeable photograph. A definition of an upgradeable image has yet to be agreed and therefore the numbers of patients in this category who will require slit-lamp biomcroscopy with Volk lens is yet to be quantified but it may be around 10 per cent. Quality assurance considerations mean that it is unlikely that all optometrists in an area will grade using slit-lamp bio. It is likely that these patients will be sent to a small number of optometrists who will be further trained and accredited to carry out slit-lamp bio on these patients.

TRAINING
It is clear that training must be competency-based and a large number of optometrists have already undergone accredited training to be able to participate in local schemes. However, it is equally clear that optometrists and ophthalmologists already have many of the competencies by virtue of the professional training they undertake. Non-ophthalmic technician graders will require even the most basic training. Competency-based training should take account of existing skills.
It seems a waste of resources for PCTs to dismantle existing successful optometric schemes, thus losing existing skills and expertise, in order to train a completely new group of personnel. In not dismantling current schemes, it is recognised that there may nevertheless be a reduction in the numbers of optometric practices involved, in order to enable audit and other criteria to be met.
THE WAY FORWARD
The establishment of a national screening programme represents a real challenge for optometry. This is an area in primary care which falls naturally within our scope of practice. It is one for which we are fully trained and in which we are best suited to provide care for patients in a way which is accessible, convenient, and is both clinically effective and cost effective.
In the near future, a matter of weeks, the advisory group will produce guidance to strategic health authorities and PCTs on the essential elements in developing a scheme. This is a significant document and it is likely that SHAs and PCTs around the country will use it as a checklist for introducing programmes in their areas.
LOCs will have to act soon to engage with PCTs and other local commissioners to ensure they are actively involved in the planning process in their areas and in all stages of the decision making process.

REFERENCES
1 Hulme SA, Tin-U A, Hardy KJ, Joyce PW. Evaluation of a district-wide screening programme for diabetic retinopathy utilizing trained optometrists using slit-lamp and Volk lenses. Diab Medicine, 2002; 19:741-745.
2 Taylor L, Riley A. Photographic study of diabetic patients. Optometry Today, 2001; 19 October.

Richard Broughton is past president of the College of Optometrists; Trevor Warburton is chairman of the AOP Professional Services Committee; and David Cartwright is vice president of the College of OptometristsFollowing publication of the implementation phase of the National Service Framework for diabetes, the National Screening Committee has established a Diabetic Retinopathy Screening Project Advisory Group to consider how a systematic national retinopathy screening programme will be introduced in England.
The Diabetic Retinopathy Screening Project Advisory Group was set up in 2002 with a number of terms of reference, including:

To advise the UK National Screening Committee on the implementation, development, review and modification of a national screening programme in England
To support local delivery to achieve the objective in line with the Diabetes National Service Framework delivery strategy that by March 2006, 'primary care trusts will have ensured that a minimum of 80 per cent of people with diabetes are offered screening for the early detection (and treatment if needed) of diabetic retinopathy as part of a systematic programme that meets the national standards, rising to 100 per cent coverage of those at risk of retinopathy by end 2007'.

Membership of the group covers a wide spectrum of professionals who are involved in diabetic retinopathy screening. The main work of the group has been delegated to three sub groups which have considered specific areas:

Education and training
Camera specification and IT
Grading, referral and quality assurance.

Optometry is represented on the main group by Richard Broughton and David Cartwright, with Trevor Warburton, Susan Blakeney and Geoff Roberson taking an active part in the sub groups.
There are a number of complex issues within the discussions that will have an influence on optometric participation. This article sets out the case that has been presented on behalf of optometry, the background to the ongoing discussions and finally the anticipated next steps that will affect optometry.
At the Project Advisory Group's most recent meeting in September, the optometric representatives gave a detailed presentation setting out clearly all the benefits that optometrist involvement will contribute to a national screening programme for diabetic retinopathy and how optometry can meet the proposed service objectives for quality assurance set out in the National Screening Committee's consultation document.
The principal points made in the submission to the advisory group were as follows:

In many parts of the UK, successful and effective co-management schemes have been established involving the participation of optometrists. Indeed, the majority of existing diabetic retinopathy screening services across the country involve optometrists
Clinical audits of a number of well-organised schemes have shown that optometrists trained to appropriate standards can safely and reliably refer sight-threatening eye disease to acceptable levels of sensitivity and specificity.

As an example, an audit of the East Suffolk Diabetic Eye Screening Scheme for the year April 2000 to March 2001 showed the following results:

84 per cent of the target population were screened during the year (3,268 individuals)
91 per cent of GP practices participated
96 per cent of patients had no, or minimal, background retinopathy
4 per cent (130 individuals) were referred to the hospital, of whom 49 were discharged, 76 were followed up at the hospital and 19 had laser treatment
150 patients were noted as having other ophthalmic conditions, including cataract and glaucoma (4.6 per cent).

Hulme et al in their evaluation of a district-wide screening programme for diabetic retinopathy utilising trained optometrists using slit-lamp and Volk lenses1 assessed sensitivity and specificity for an optometric scheme:

Sensitivity for STED was 87 per cent
Specificity was 91 per cent.

A guidance document on diabetic co-management has been produced by the College of Optometrists, the Royal College of Ophthalmologists, the Royal College of General Practitioners, the Royal College of Physicians and Diabetes UK (formerly the British Diabetic Association), with the support of the Department of Health. Thus it can be seen that the involvement of optometrists in the care of diabetics is well supported by all the relevant professional and special interest groups.
Feedback indicates that where successful optometric schemes are in place, there is a desire to continue those schemes, simply building in the use of digital cameras. Examples of such schemes include a business-specific model from a group of PCTs in Greater Manchester and the well-established Dorset optometric scheme.

BENEFITS OF OPTOMETRIC SCHEMES
A National Screening Committee consultation paper has proposed standards for developing quality assurance and lists 17 service objectives. Optometrist-based schemes will ensure that many of these are met in an appropriate manner and will best meet the needs of patients.
The following gives the significant service standards relating to optometric schemes and the main points that highlight the benefits of optometric involvement.

To maximise the number of invited persons accepting the test (Service objective 4). Optometrists' practices are conveniently located and easily accessible, and can offer flexible appointment times. Good accessibility will increase the percentage uptake of patients within the schemes. Together with the consequent reduction in travel time and costs for patients, this will encourage better compliance. Optometrists' practices are situated in almost all communities and there are sufficient optometrists available to provide screening services for all patients with diabetes within their local vicinity.
To ensure grading is accurate (6): There is concern to avoid tiredness leading to errors. Major causes of errors are monotony and boredom due to a repetitive workload. The mixture of optometrists' workload, which encompasses time allocated to involvement in formal screening schemes and routine optometric practice, adds interest rather than the reverse, and avoids both these elements.
To ensure timely reporting of abnormal screening results (8): Optometrists have an excellent knowledge of normal retinal appearances through their everyday work in examining large numbers of patients. Accredited optometrists therefore have the skills to grade as well as screen. Thus there is no delay between screening and grading, and the timely referral of appropriate cases is ensured. The proportion of ungradeable images could be as high as 10 per cent, depending on the definition of an ungradeable image. Almost all of the ungradeable images are from patients over 70 years of age2 and this age group should not be subject to avoidable extra appointments. The NSC will need to define what constitutes an ungradeable image for the national programme.
To ensure both GP and patient are informed of negative results (11). Optometrists are able to screen and grade at the same appointment and can therefore give the patient the results immediately, whether negative or otherwise, thus avoiding unnecessary anxiety on the part of the patient.
To minimise the anxiety associated with screening (15). See comment for (11) above.
To ensure timely rescreening (16). Optometric practices have short waiting lists and are used to operating their own recall systems, and could thus easily fit into a central recall system. Optometry can also be very flexible as to capacity, with the ability to cope with the gradually rising demand for screening over the coming years.

COMPARATIVE COSTS
Optometrist-based schemes can offer an extremely cost-effective resource to help meet the proposed service objectives as set out in the National Screening Committee's consultation paper.
It has been suggested that optometry schemes are more expensive than centralised camera schemes. However, in making this claim it does not seem to be the case that truly comparative costs are used and it is often the case that the budgets quoted for alternative schemes do not include realistic infrastructure overheads.
There is much speculation that mobile or centralised screening will be very cheap, but little evidence to support this. In 2000, in Rochdale, 4,800 patients were screened by digital camera at a cost of £126,000 in staff (£26.25 per event), plus capital costs and no account of accommodation costs.
A figure around this sum may well be appropriate for an optometric camera-based scheme, but would be likely to include overhead costs as well as screening and primary grading. The optometry fee could also be agreed as a 'per screening fee', so that failure to attend does not add to PCT costs. Some simplistic calculations, which appear to show a low cost of centralised or mobile screening, are actually a 'cost per person with diabetes' which make no allowance for 'DNAs' ('did not attend').
The costs to the patient in terms of time and travel should also not be overlooked. Many will find it more convenient to be screening locally, and the NSC itself said that 'Épatients normally managed in primary care should not necessarily be expected to travel long distances to the diabetes centre just to be screened'.
In comparing costs it is tempting to ignore the cost of buildings and overheads when an existing location is chosen. Health economists will say that this is incorrect and like must always be compared with like, so the overheads of an optometric practice must be compared with the overheads of a hospital or primary care resource centre.

AUDIT
The National Screening Committee regards quality assurance as a key issue in ensuring that the screening programme delivers a high quality service. There is a particular concern that problems encountered in screening for breast and cervical cancer should not be repeated in retinopathy screening. An important issue is the minimum numbers of patients graders will need to see to ensure statistically sound quality assurance (QA). The evidence base for deciding on appropriate minimum numbers is very limited and the initial proposals requiring graders to see a minimum of 1,000 patients a year do not appear to be based on specific evidence and if adopted would effectively exclude optometry from providing a screening service which would include primary grading.
While it is accepted that there is a requirement for quality assurance to be statistically robust, optometry has argued that it is also important that numbers are realistically achievable and do not exclude from participation well-trained professionals with the appropriate skills. Numbers should be set at realistic and pragmatic levels.
This topic has generated more discussion and debate than any other in the project advisory group. It has now been recognised that there need to be specific standards for optometric schemes. The current proposal is that an optometrist involved in screening patients within their own practice, as part of a scheme, will photograph and primary grade Ð that is determine the presence or absence of retinopathy and the presence or absence of referable retinopathy. They should see a minimum of 250 patients and if they see less than 500, they should have additional patient screening episodes, sent electronically, to grade.
For example, if the optometrist sees 250 patients they will have another 250 sets of patient images to grade. However, it is not yet agreed how many additional electronic gradings an optometrist who examines between 250 and 500 patients will have to see. The quality assurance then will be based on reviewing a proportion of decisions made on a total of 2,000 images graded (500 patient episodes, times four pictures per patient).
A significant minority of patients who attend for screening by digital photography will produce an upgradeable photograph. A definition of an upgradeable image has yet to be agreed and therefore the numbers of patients in this category who will require slit-lamp biomcroscopy with Volk lens is yet to be quantified but it may be around 10 per cent. Quality assurance considerations mean that it is unlikely that all optometrists in an area will grade using slit-lamp bio. It is likely that these patients will be sent to a small number of optometrists who will be further trained and accredited to carry out slit-lamp bio on these patients.

TRAINING
It is clear that training must be competency-based and a large number of optometrists have already undergone accredited training to be able to participate in local schemes. However, it is equally clear that optometrists and ophthalmologists already have many of the competencies by virtue of the professional training they undertake. Non-ophthalmic technician graders will require even the most basic training. Competency-based training should take account of existing skills.
It seems a waste of resources for PCTs to dismantle existing successful optometric schemes, thus losing existing skills and expertise, in order to train a completely new group of personnel. In not dismantling current schemes, it is recognised that there may nevertheless be a reduction in the numbers of optometric practices involved, in order to enable audit and other criteria to be met.
THE WAY FORWARD
The establishment of a national screening programme represents a real challenge for optometry. This is an area in primary care which falls naturally within our scope of practice. It is one for which we are fully trained and in which we are best suited to provide care for patients in a way which is accessible, convenient, and is both clinically effective and cost effective.
In the near future, a matter of weeks, the advisory group will produce guidance to strategic health authorities and PCTs on the essential elements in developing a scheme. This is a significant document and it is likely that SHAs and PCTs around the country will use it as a checklist for introducing programmes in their areas.
LOCs will have to act soon to engage with PCTs and other local commissioners to ensure they are actively involved in the planning process in their areas and in all stages of the decision making process.

REFERENCES
1 Hulme SA, Tin-U A, Hardy KJ, Joyce PW. Evaluation of a district-wide screening programme for diabetic retinopathy utilizing trained optometrists using slit-lamp and Volk lenses. Diab Medicine, 2002; 19:741-745.
2 Taylor L, Riley A. Photographic study of diabetic patients. Optometry Today, 2001; 19 October.

Richard Broughton is past president of the College of Optometrists; Trevor Warburton is chairman of the AOP Professional Services Committee; and David Cartwright is vice president of the College of Optometrists