Diabetes is a major health problem facing health care providers worldwide. 4% of New Zealanders have diabetes with this number increasing as the population ages and obesity rates increase.1 Among these patients 27% have some degree of diabetic retinopathy, with 12% severe enough to need care from an ophthalmologist.2 With rates of diabetes rising the numbers suffering from diabetic eye disease are rising, providing a challenge for ophthalmologists, and particularly the public health system.Screening for diabetic eye disease is vital to reduce the burden of disease, however it is an expensive process. Appropriate screening systems combined with timely treatment however make them cost effective in terms of sight years preserved.3-6 To achieve this collaboration with other ophthalmic healthcare professionals important in reducing the burden of care on the public health system.In 2002 the WIPA (Wellington Independent Practitioners Association) was established to provide diabetic eye screening services for the Wellington region. Screening involves non-mydriatic fundus photographs, with mydriasis used if an unsatisfactory image is obtained, and slit lamp biomicroscopy performed by the optometrist if this is unsatisfactory. Photographs are graded according to the National Diabetes Retinal Screening Grading System with grades for both maculopathy and retinopathy. Maculopathy grading is from M0 (least severe) to M5 (most severe), in addition there is a category MT for stable treated maculopathy.7Patients graded M3 or higher, or those with other pathology are recommended to be referred to an ophthalmologist for treatment. The criteria for M3 maculopathy are exudates or thickening within two disc diameters, but outside one disc diameter from the centre of the macula, and for M4 maculopathy exudates or retinal thickening within one disc diameter of the centre of the macula, with no foveal involvement.7Furthermore those who are unable to have a satisfactory photograph or retinal examination by the screening optometrist due to reasons such as cataract are referred for further management. In addition to diabetic eye disease screening detects significant amounts of incidental non-diabetic ocular disease, resulting in improved visual outcomes for these patients.5,8,9Grading of photographs has been an issue for diabetic screening programmes worldwide. Sallam et al found only 72.3% agreement with photographic grading and hospital slit lamp biomicroscopy grading.10 Benbasat and Pollack found variable rates of inter- and intra-examiner agreement when grading retinal photographs. Counts of microaneurysms were found to be in agreement only 39 to 85% of the time with two assessments by the same examiner.11 This has lead to other methods, including development of automatic grading by computer systems. Automated computer grading has been shown to be safe and reduce the burden on healthcare professionals.5,12,13 Some studies even found computerised grading more sensitive in detection of haemorrhages and exudates.14 Despite this the ability to detect other pathology, and the potential to miss other serious conditions such as tumours make it less attractive.8,15,16With the need for diabetic screening established, it is less clear how best to carry out this screening, and in collaborative programmes with other ocular healthcare professionals at what point referral for review by an ophthalmologist is required. This paper reviews the rates of progression of diabetic maculopathy from those graded M3 initially by optometrists, and assesses the safety of the current referral criteria referring M4 but not M3 used by the WIPA Screening Programme.Methods From the Wellington Independent Practitioners Association (WIPA) screening database all patients with M3 maculopathy (as defined by the New Zealand National Diabetes Retinal Screening Grading System 2006) were identified from the inception of the database in 2002 to 30 June 2009. M3 or mild maculopathy was defined as exudates and/or retinal thickening within 2 disc diameters of the centre of the macula but outside 1 disc diameter. 45 eyes of 54 patients enrolled in the WIPA Screening Programme were studied. Photographic screening data was collated from the initial presentation with M3 maculopathy, and their first subsequent visit. The photographs for each visit were graded by the grading optometrist and a consultant ophthalmologist. Those photographs graded M3 at the initial visit by both the optometrist and ophthalmologist were included. Exclusion criteria were those patients who had been referred to clinics for pathology in the fellow eye, had been lost to follow-up, had no photos available or for whom the grading deferred between the optometrist and ophthalmologist. Main outcome measures were the percentage of patients whose maculopathy progressed to greater than M3 at the subsequent visit, the duration between screening visits, and the number of patients with a significant reduction in Snellen visual acuity (more than one line) among those progressing from M3 maculopathy. Results Of the 54 eyes graded M3 by both the screening optometrist and consultant ophthalmologist, 18 or 33.3% progressed to M4 maculopathy at the subsequent visit. 3.7% or two of 54 eyes improved from M3 to M2 maculopathy, and 63.0% or 34 of 54 eyes remained M3 (Figure 1). Mean duration between screening photographs was 255.0 (\u00b1 59.5) days. Despite progressing from M3 to M4, only one patient (5.6%) had a reduction of vision by more than one line of Snellen acuity at follow up. This was not due to progression of their diabetic maculopathy. Figure 1. Rates of progression of maculopathy Discussion Photographic screening for diabetes enables patients to see the photographs and discuss them, and rates of attendance improve when they are taken and screening locations spread about the community. This results in further opportunities to ephasise the importance of diabetic control and taking ownership of ones health outcomes. Health professionals such as optometrists are more effective at this than a photographer in a technician role, and may help improve not only ocular outcomes, but also more general health outcomes. The rates of progression of maculopathy from M3 to M4 were of clinical significance. Despite this no patient had worsening visual acuity at the subsequent visit due to the diabetic eye disease. This suggests patients with M3 maculopathy are safe to remain under a well managed screening programme with accurate grading, clear referral pathways for those who progress, and a peer review system. This guideline would be consistent with the British guidelines, which recognise only maculopathy with exudates or retinal thickening within 1 disc diameter of the central macula as requiring referral.15 For the Wellington Screening Programme this means appropriate training for graders and supervision of the screening optometrists by an Ophthalmologist is vital. Finally the mean interval between screening photographs at 255 days was slightly outside the guideline set of 6 months (183 days). As with the issue of when a referral for review by an ophthalmologist is required, the appropriate interval between screening visits has not been clearly established 7. This study found rates of progression for those with moderate maculopathy to be significant, however this may have been reduced if the screening interval were shorter. Despite this no patient lost vision as a result of this progression, suggesting an interval of 255 days may be safe. While providing useful information there is more work to be done before it can be stated confidently that patients with M3 maculopathy are safe to remain under a screening programme. Firstly this was a relatively small sample of 54 eyes, with only 18 progressing to M4 maculopathy. Furthermore there were a number of people excluded for non-agreement between the grading optometrist and ophthalmologist, non-attendance, missing photographs and other reasons. It is possible non-attenders are at higher risk of visual loss, possibly therefore requiring referral at an earlier stage. While these issues may influence the progression and interval sections there is no reason to suggest the grading accuracy would be any different among these groups. These results suggest if grading is sufficiently accurate those with M3 maculopathy are safe to remain under an optometrist run screening programme at an interval of 255 days between visits. Despite this relative safety, with progression rates at 33.3% a new grading system to help identify patients at increased risk of progression may be useful if patients with M3 maculopathy were to remain under the care of a primary screener or optometrist.

Diabetes is a major health problem facing health care providers worldwide. 4% of New Zealanders have diabetes with this number increasing as the population ages and obesity rates increase.1 Among these patients 27% have some degree of diabetic retinopathy, with 12% severe enough to need care from an ophthalmologist.2 With rates of diabetes rising the numbers suffering from diabetic eye disease are rising, providing a challenge for ophthalmologists, and particularly the public health system.Screening for diabetic eye disease is vital to reduce the burden of disease, however it is an expensive process. Appropriate screening systems combined with timely treatment however make them cost effective in terms of sight years preserved.3-6 To achieve this collaboration with other ophthalmic healthcare professionals important in reducing the burden of care on the public health system.In 2002 the WIPA (Wellington Independent Practitioners Association) was established to provide diabetic eye screening services for the Wellington region. Screening involves non-mydriatic fundus photographs, with mydriasis used if an unsatisfactory image is obtained, and slit lamp biomicroscopy performed by the optometrist if this is unsatisfactory. Photographs are graded according to the National Diabetes Retinal Screening Grading System with grades for both maculopathy and retinopathy. Maculopathy grading is from M0 (least severe) to M5 (most severe), in addition there is a category MT for stable treated maculopathy.7Patients graded M3 or higher, or those with other pathology are recommended to be referred to an ophthalmologist for treatment. The criteria for M3 maculopathy are exudates or thickening within two disc diameters, but outside one disc diameter from the centre of the macula, and for M4 maculopathy exudates or retinal thickening within one disc diameter of the centre of the macula, with no foveal involvement.7Furthermore those who are unable to have a satisfactory photograph or retinal examination by the screening optometrist due to reasons such as cataract are referred for further management. In addition to diabetic eye disease screening detects significant amounts of incidental non-diabetic ocular disease, resulting in improved visual outcomes for these patients.5,8,9Grading of photographs has been an issue for diabetic screening programmes worldwide. Sallam et al found only 72.3% agreement with photographic grading and hospital slit lamp biomicroscopy grading.10 Benbasat and Pollack found variable rates of inter- and intra-examiner agreement when grading retinal photographs. Counts of microaneurysms were found to be in agreement only 39 to 85% of the time with two assessments by the same examiner.11 This has lead to other methods, including development of automatic grading by computer systems. Automated computer grading has been shown to be safe and reduce the burden on healthcare professionals.5,12,13 Some studies even found computerised grading more sensitive in detection of haemorrhages and exudates.14 Despite this the ability to detect other pathology, and the potential to miss other serious conditions such as tumours make it less attractive.8,15,16With the need for diabetic screening established, it is less clear how best to carry out this screening, and in collaborative programmes with other ocular healthcare professionals at what point referral for review by an ophthalmologist is required. This paper reviews the rates of progression of diabetic maculopathy from those graded M3 initially by optometrists, and assesses the safety of the current referral criteria referring M4 but not M3 used by the WIPA Screening Programme.Methods From the Wellington Independent Practitioners Association (WIPA) screening database all patients with M3 maculopathy (as defined by the New Zealand National Diabetes Retinal Screening Grading System 2006) were identified from the inception of the database in 2002 to 30 June 2009. M3 or mild maculopathy was defined as exudates and/or retinal thickening within 2 disc diameters of the centre of the macula but outside 1 disc diameter. 45 eyes of 54 patients enrolled in the WIPA Screening Programme were studied. Photographic screening data was collated from the initial presentation with M3 maculopathy, and their first subsequent visit. The photographs for each visit were graded by the grading optometrist and a consultant ophthalmologist. Those photographs graded M3 at the initial visit by both the optometrist and ophthalmologist were included. Exclusion criteria were those patients who had been referred to clinics for pathology in the fellow eye, had been lost to follow-up, had no photos available or for whom the grading deferred between the optometrist and ophthalmologist. Main outcome measures were the percentage of patients whose maculopathy progressed to greater than M3 at the subsequent visit, the duration between screening visits, and the number of patients with a significant reduction in Snellen visual acuity (more than one line) among those progressing from M3 maculopathy. Results Of the 54 eyes graded M3 by both the screening optometrist and consultant ophthalmologist, 18 or 33.3% progressed to M4 maculopathy at the subsequent visit. 3.7% or two of 54 eyes improved from M3 to M2 maculopathy, and 63.0% or 34 of 54 eyes remained M3 (Figure 1). Mean duration between screening photographs was 255.0 (\u00b1 59.5) days. Despite progressing from M3 to M4, only one patient (5.6%) had a reduction of vision by more than one line of Snellen acuity at follow up. This was not due to progression of their diabetic maculopathy. Figure 1. Rates of progression of maculopathy Discussion Photographic screening for diabetes enables patients to see the photographs and discuss them, and rates of attendance improve when they are taken and screening locations spread about the community. This results in further opportunities to ephasise the importance of diabetic control and taking ownership of ones health outcomes. Health professionals such as optometrists are more effective at this than a photographer in a technician role, and may help improve not only ocular outcomes, but also more general health outcomes. The rates of progression of maculopathy from M3 to M4 were of clinical significance. Despite this no patient had worsening visual acuity at the subsequent visit due to the diabetic eye disease. This suggests patients with M3 maculopathy are safe to remain under a well managed screening programme with accurate grading, clear referral pathways for those who progress, and a peer review system. This guideline would be consistent with the British guidelines, which recognise only maculopathy with exudates or retinal thickening within 1 disc diameter of the central macula as requiring referral.15 For the Wellington Screening Programme this means appropriate training for graders and supervision of the screening optometrists by an Ophthalmologist is vital. Finally the mean interval between screening photographs at 255 days was slightly outside the guideline set of 6 months (183 days). As with the issue of when a referral for review by an ophthalmologist is required, the appropriate interval between screening visits has not been clearly established 7. This study found rates of progression for those with moderate maculopathy to be significant, however this may have been reduced if the screening interval were shorter. Despite this no patient lost vision as a result of this progression, suggesting an interval of 255 days may be safe. While providing useful information there is more work to be done before it can be stated confidently that patients with M3 maculopathy are safe to remain under a screening programme. Firstly this was a relatively small sample of 54 eyes, with only 18 progressing to M4 maculopathy. Furthermore there were a number of people excluded for non-agreement between the grading optometrist and ophthalmologist, non-attendance, missing photographs and other reasons. It is possible non-attenders are at higher risk of visual loss, possibly therefore requiring referral at an earlier stage. While these issues may influence the progression and interval sections there is no reason to suggest the grading accuracy would be any different among these groups. These results suggest if grading is sufficiently accurate those with M3 maculopathy are safe to remain under an optometrist run screening programme at an interval of 255 days between visits. Despite this relative safety, with progression rates at 33.3% a new grading system to help identify patients at increased risk of progression may be useful if patients with M3 maculopathy were to remain under the care of a primary screener or optometrist.