You are here

Article Text

Article menu

Download PDFPDF

Original articles
Clinical science
Rapid assessment of avoidable blindness and diabetic retinopathy in Taif, Saudi Arabia
  1. Abdul Hamid Al Ghamdi1,
  2. Mansur Rabiu2,
  3. Saad Hajar3,
  4. David Yorston4,
  5. Hannah Kuper5,
  6. Sarah Polack5
  1. 1Department of Ophthalmology, Faculty of Medicine, Taif University, Taif, Saudi Arabia
  2. 2International Agency of Prevention of Blindness, Riyadh, Saudi Arabia
  3. 3King Fahad Medical City, Riyadh, Saudi Arabia
  4. 4Tennent Institute of Ophthalmology, Glasgow, UK
  5. 5Department of Clinical Research, London School of Hygiene & Tropical Medicine, London, UK
  1. Correspondence to Hannah Kuper, Department of Clinical Research, London School of Hygiene & Tropical Medicine, London WC1E 7HT, UK; hannah.kuper{at}lshtm.ac.uk

Abstract

Background/aims To estimate the prevalence of blindness, diabetes and diabetic retinopathy (DR) in Taif, Saudi Arabia using the Rapid Assessment of Avoidable Blindness (RAAB) framework.

Methods 66 clusters of 50 people aged ≥50 years were randomly selected. Participants underwent visual acuity measurement and examination by an ophthalmologist. DR among diabetic participants (previous diagnosis and/or random blood glucose >200 mg/dl) was assessed through dilated fundus examination by an ophthalmologist using a direct and indirect ophthalmoscope (‘clinical examination’) and dilated digital fundus photographs graded by a retinal specialist following the Scottish DR grading system (‘reference standard’).

Results 3052 (93%) out of 3300 eligible people were examined. The prevalence of blindness was 2.6% (95% CI 2.0% to 3.2%). Posterior segment diseases (44%) and cataract (41%) were the leading causes of blindness. The estimated prevalence of diabetes was 29.7% (28.1% to 31.4%), among whom the prevalence of DR was 36.8% (33.3% to 40.2%) and sight-threatening DR (STDR) was 17.5% (CI 15.1% to 20.0%). Agreement was good (κ>0.6) between the clinical examination and reference standard for any DR and STDR.

Conclusions There was a high prevalence of diabetes, DR and STDR. It was possible to assess diabetes and DR within RAAB but it increased the survey duration, cost and complexity.

  • Diabetic retinopathy
  • Saudi Arabia
  • blindness
  • survey

https://doi.org/10.1136/bjophthalmol-2012-301874

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    The WHO estimates that there are 39 million blind people in the world, and another 246 million are visually impaired.1 The magnitude and causes of vision loss vary considerably both between and within countries, and reliable up-to-date data are needed to develop locally appropriate blindness prevention programmes. A national eye survey in Saudi Arabia conducted in the 1980s estimated the all-age prevalence of blindness to be 1.5%.2 Cataract and trachoma were the leading causes. Since then, the country has experienced substantial improvements in socioeconomic indices, with a consequent increase in life expectancy and shift from infectious to chronic diseases as the leading causes of death.3 These transitions are likely to be reflected in changes in the epidemiology of eye diseases, including reductions in preventable and easily treatable conditions, but increased challenges from diseases related to lifestyle and ageing (eg, diabetic retinopathy—DR). However, data documenting the current situation in Saudi Arabia are lacking.

    The prevalence of diabetes is rapidly increasing worldwide and particularly in the Middle East.4 Saudi Arabia is among the top 10 countries with the highest prevalence of diabetes4; a national survey conducted between 1999 and 2005 found nearly a quarter of people aged 30–70 years had diabetes.5 DR is responsible for 3.9% of blindness globally,6 but this proportion is likely to be greater in areas with a high prevalence of diabetes, and will grow as diabetes increases, populations age and other conditions (eg, cataract) are brought under control. Population-based data on the prevalence of DR are sparse, particularly in low- and middle-income settings, because these data are thought to be difficult and expensive to obtain. Yet data on the prevalence of DR in Saudi Arabia are urgently needed to inform the appropriate planning of DR services.

    The Rapid Assessment of Avoidable Blindness (RAAB) is a population-based survey of visual impairment and eye care services among people aged 50 years and over.7 RAAB is undertaken to estimate the prevalence of blindness and visual impairment and its causes in a specific geographical area, usually at the district or province level. RAAB examination methodology relies on visual acuity assessment and assessment with a direct ophthalmoscope, and so, focuses primarily on the prevalence of avoidable blindness, that is, due to cataract, refractive errors, trachoma, onchocerciasis and other corneal scarring. A new method has been developed to estimate diabetes and DR prevalence within RAAB (RAAB+DR).8 This method needs to be tested in different settings in terms of its feasibility and reliability.

    The aim of this study is to estimate the prevalence of blindness, diabetes and DR in Taif, Saudi Arabia, using the RAAB+DR methodology.

    Materials and methods

    RAAB

    The survey was undertaken in the Taif region of Saudi Arabia. A sample size of 3300 was required, given an expected blindness prevalence of 3% among people aged ≥50 years,9 a precision of 25%, a confidence of 95%, design effect of 1.5 and 10% non-response.

    Clusters were selected through probability proportion to size sampling. Within clusters, households were selected using compact segment sampling: clusters were divided into segments containing approximately 50 people aged ≥50 years, one segment was randomly chosen, and survey teams visited households door-to-door until 50 eligible people were included. Absentees were revisited twice more, and if still unavailable, details on likely vision status were obtained from relatives or neighbours.

    At the household, participants underwent a visual acuity (VA) examination in daylight using a tumbling ‘E’ chart. Pinhole vision was measured for eyes with VA<6/18. The lens status of all eyes was assessed by an ophthalmologist using a torch or direct ophthalmoscope in a shaded area. Eyes with VA <6/18 were examined to determine the main cause of visual impairment.

    Diabetes and DR

    A finger-prick random blood glucose (RBG) test was undertaken by a nurse using a digital glucometer. Diabetes was defined as either a self-report of previous diagnosis of diabetes, self-report of taking treatment to control blood glucose or an RBG level >200 mg/dl.10

    Diabetic participants underwent a dilated fundus examination at the household using a direct and indirect ophthalmoscope with a 20 dioptre lens (‘clinical examination’). DR was classified using the Scottish DR grading system (table 1).11 In addition, diabetic participants were referred to a local centre on the same day for a digital fundus photograph. Using a Topcon non-mydriatic retinal camera NW300, a single 45X dilated retinal image was taken which centred on the fovea and included the optic disc and temporal vascular arcades following the Scottish DR grading protocol. The images were graded in the UK by a retinal specialist with extensive experience, using the Scottish DR grading scheme (DY) to provide a ‘reference standard’ for comparison with clinical examination.

    View this table:
    Table 1

    Scottish diabetic retinopathy grading scheme11

    Survey teams

    There were four survey teams consisting of one ophthalmologist and two nurses. Two teams completed one cluster together daily, rather than one team per cluster per day as recommended in a standard RAAB. Two trained optometrists were responsible for taking the fundus photographs. Training was conducted over 8 days and included assessment of the interobserver agreement for vision, lens assessment, causes of visual impairment and DR grading.

    Data analysis

    Data on prevalence and causes of blindness were analysed using the automated RAAB software. We used STATA V.11 to estimate the prevalence of diabetes and DR with the SVY command to account for the cluster design. Logistic regression analyses were undertaken to assess the association between (a) diabetes, (b) vision loss and (c) DR with age, gender, education, nationality and RBG level. We compared the DR grading from the clinical examination and retinal photograph grading and calculated κ values, sensitivity, specificity, positive and negative values for the presence of DR and grades of DR.

    Ethical considerations

    Ethical approval was granted by the Ethics Committees of the London School of Hygiene & Tropical Medicine, the King Fahad Medical City Riyadh, and the Ministry of Health, Kingdom of Saudi Arabia. Written/thumbprinted consents were obtained from the participants. People identified with new or uncontrolled diabetes and/or ocular conditions requiring further investigation, treatment or follow-up were referred as appropriate.

    Results

    Socio-demographic characteristics and diabetes

    Out of 3300 eligible people, 3052 (93%) were examined, 153 (5%) were not available, 65 (2%) refused to participate and 30 (1%) were unable to communicate. The age and gender distribution of respondents (63.3 years, 54% males) and non-respondents (61.2 years, 53% males) were similar.

    Diabetes status was assessed in 63 out of the 66 clusters. Of the 2917 (93%) who were examined in these clusters, 51 people without a prior diabetes diagnosis refused RBG assessment and were, therefore, excluded from the diabetes and DR analyses.

    The prevalence of diabetes was 29.7% (95% CI 28.1% to 31.4%). Diabetes was more common among people aged 60–69 years (OR 1.76, 95% CI 1.45 to 2.16) and 70–79 years (OR 2.16, 1.72 to 2.71) compared with people aged 50–59 years, but was not significantly different among people aged ≥80 years. Saudi Arabian nationals were more likely to have diabetes compared with those of other nationalities (OR 2.28, 95% CI 1.72 to 3.02). These associations remained significant with multivariate adjustment for age, gender, education and nationality (data not presented). There was no association between diabetes and gender or education in the univariate analysis, although higher education became associated with diabetes after multivariate adjustment (OR 95% CI 1.49, 1.05 to 2.12).

    Most people with diabetes (88%) had a previous diagnosis of diabetes. More than half (56%) the people with known diabetes had poor control (RBG >200 mg/dl).

    Blindness

    Seventy-nine people were blind (presenting bilateral VA <3/60) giving a prevalence of blindness of 2.6% (95% CI 2.0 to 3.2, (table 2). The prevalence of bilateral severe visual impairment (SVI, VA<6/60—≥3/60) was 1.7% (95% CI 1.3% to 2.2%) and visual impairment (VI, VA<6/18—≥6/60) was 7.5% (95% CI 6.5% to 8.5%). Extrapolating to the age and gender distribution of Taif, there are an estimated 11 000 people aged ≥50 years with bilateral blindness, 8400 with SVI and 39 200 with VI.

    View this table:
    Table 2

    Prevalence of blindness, severe visual impairment and visual impairment by eyes and person (presenting vision) in Taif

    Bilateral vision loss (VA<6/18) was more common among older people (p-for-trend <0.001), women (OR 1.5 95% CI 1.2 to 1.9), participants with no formal education (OR 3.9, 95% CI 2.8 to 5.4) and Saudi nationals (OR 2.4, 95% 1.5 to 3.7). The prevalence of VI, SVI and blindness was higher among people with diabetes compared with people without diabetes (6.4%, 1.4% and 2.3% compared with 10.1%, 2.3% and 3.3%, respectively), although this difference was only statistically significant for the VI group (OR 1.45, 95% CI 1.08 to 1.94).

    Causes of blindness

    Posterior segment diseases were the leading cause of bilateral blindness (44%, table 3). This included DR which was responsible for 10% of blindness, 15% SVI and 8% VI. Cataract was the leading single cause of blindness (41%) and SVI (45%) and was responsible for 35% of VI. More than a third of VI was due to refractive error. Over half (56%) of all blindness was ‘avoidable’, while 72% of SVI and 84% of VI was avoidable.

    View this table:
    Table 3

    Principal cause of blindness, severe visual impairment and visual impairment in persons (presenting vision) in Taif

    Among people without diabetes, cataract was the leading cause of blindness (40%) followed by other posterior segment diseases (38%). For people with diabetes, 29% of blindness was due to DR, 32% to other posterior segment diseases, and 32% to cataract.

    Diabetic retinopathy

    Of the 852 participants with diabetes, 612 (72%) had a dilated eye examination. For 16 (3%) participants, eyes could not be graded. Participants who did not have a fundus examination were more likely to be male (OR 1.5 95% CI 1.1 to 2.0), but were otherwise similar to those examined.

    Approximately one-third of people with diabetes had signs of retinopathy (34.5% 95% CI 31.0 to 38.1), and a fifth had maculopathy (20.3%, 17.4 to 23.2, table 4). The prevalence of any retinopathy and/or maculopathy was 36.8% (33.3 to 40.2) and sight-threatening DR (STDR) was 17.5% (15.1 to 20.0). Of those with STDR, 27% had evidence of previous laser treatment. There was a higher prevalence of any retinopathy and/or maculopathy among people with known diabetes (37.7%, 34.0 to 41.4) compared with newly diagnosed diabetes (28.1%, 15.0 to 41.1), but this difference was not significant (p=0.13). The prevalence of any DR across all participants (ie, including non-diabetic) was 7.9% (6.8 to 8.9). Extrapolating to the population of Taif, it is expected that there are approximately 26 000 people aged ≥50 years with diabetes, 9000 with DR and 4500 with STDR.

    View this table:
    Table 4

    Prevalence of diabetic retinopathy (DR) among diabetic participants and the full survey population

    DR prevalence among people with diabetes increased with increasing age (p<0.001) and RBG level (p<0.001), but was not associated with gender, education or nationality (table 5).

    View this table:
    Table 5

    Logistic regression analysis of the association between diabetic retinopathy (DR) and socio-demographic and diabetes characteristics

    Cataract

    Cataract surgery coverage was high for both persons and eyes: 90% of people and 78% of eyes needing surgery at VA<3/60 had received it, 88% and 73% respectively at VA<6/60 and 78% and 59% at VA<6/18. With available correction, 62% of eyes had good outcome after cataract surgery (VA>6/18), and this increased to 69% with pinhole correction. The proportion of eyes with borderline (VA <6/18–6/60) and poor outcome (VA<6/60) was 18% and 19%, respectively. Poor outcome was most commonly due to a concurrent cause of blindness (55%), surgery-related complications (36%) or uncorrected refractive error (12%). The most frequent barriers to surgery among people blind from cataract were ‘old age, need not felt’ (26%), ‘unaware of treatment’ (17%), ‘cannot afford the operation’ (11%) or ‘fear of losing sight’ (11%).

    Survey feasibility

    With two teams working together to complete each cluster, it was possible to include the RAAB+DR protocol, although this increased the duration, running costs, personnel and resources required. Comparing the DR diagnosis by clinical examination at the household with the retinal photographs, we generally found good agreement (κ >0.6) (table 6). Sensitivity and specificity values were also high.

    View this table:
    Table 6

    Diabetic retinopathy grading agreement (by person) between clinical examination during survey and retinal photograph graded by retinal specialist

    Discussion

    This RAAB+DR survey was the first population-based study of visual impairment conducted in Saudi Arabia in 20 years,2 ,12 and the only population-based survey of DR conducted in Saudi Arabia. Among people aged ≥50 years, the prevalence of blindness was 2.5%, and posterior segment diseases (44%) and cataract (41%) were the leading causes of blindness. Nearly a third of the survey population had diabetes (29.7%). Among people with diabetes, there was a high prevalence of DR (36%) and STDR (17%).

    Countries in the Middle East have some of the highest prevalence of diabetes in the world.4 Our estimates are broadly consistent with the national survey conducted from 1999 to 2005 (>34% among people aged 40–70 years) and a recent survey in Riyadh (46% among people aged 46–80 years), and suggest an alarmingly high burden of diabetes in this setting.

    The estimated prevalence of blindness was considerably lower than in the national eye survey conducted in the 1980s which reported a 15% prevalence of blindness in people aged >60 years.2 This may reflect improvements in eye care services and living standards. Our blindness prevalence estimates were within the range of other recent RAABs carried out in the Middle East, including Iran (1.5%),13 Qatar (1.5%),14 and the Occupied Palestinian Territories (3.4%).9 Blindness was more common among females, older participants and people lacking in formal education, all findings which have been observed elsewhere.15–17 Cataract remains a leading cause of blindness (41%) in Taif, as was also observed in Iran (32%),13 and the Occupied Palestinian Territories (55%).9 Visual outcome from cataract surgery did not meet the WHO recommended targets (80% VA >6/18 with available correction),18 indicating a need to monitor cataract outcomes to improve quality.19

    More than a third of people with diabetes had signs of DR, and approximately one in six had sight-threatening stages of the disease. DR was responsible for 10% of blindness, and this is likely to increase further as the magnitude of diabetes rises further. Laser treatment is effective in reducing loss of sight among people with STDR,20 however, most people with STDR had not previously undergone laser treatment, highlighting the need for improved detection and management. Poor glycaemic control is a well-established risk factor for diabetes complications,21 ,22 and was very common and associated with DR prevalence in the current study, suggesting that improved awareness and management of glycaemic levels should be a priority.

    We found it possible to conduct diabetes and DR assessment within RAAB in this setting, and these findings can be used to inform planning. However, it added considerable time and complexity to the survey and, therefore, RAAB+DR should only be applied in settings with sufficient time and resources available, and where high diabetes prevalence is expected. Indirect ophthalmoscopy has reasonable agreement with fundus photography in a clinical setting.23 We also found good agreement with retinal photographs graded by a retinal expert for any retinopathy, any maculopathy and STDR. This lends support to reliability of prevalence estimates derived from RAAB+DR, providing the ophthalmologists are well skilled. Use of a retinal camera may further increase sensitivity and specificity, but would add considerable logistical complications and expense.

    There were a number of limitations. Diabetes was diagnosed only using random blood glucose, and only people aged ≥50 years were included. A quarter of diabetic participants did not undergo dilated examination, although there was no difference between those who did and did not undergo examination in characteristics that were also associated with risk of DR (ie, age and RBG). We used the Scottish grading system for classification of DR. The scheme is designed around a 45° field of view captured in a single retinal photo, as a single photo is adequate for DR screening,24–26 however, some cases may be missed when using this method.

    In summary, using the new RAAB+DR methodology, this survey found a high prevalence of diabetes and DR among people aged ≥50 years in Taif, Saudi Arabia. Nearly a fifth of people with diabetes had STDR, of which the majority had not received laser treatment.

    Acknowledgments

    We thank all the survey team members including the ophthalmologists (Azza Asal, Idris Agbabiaka, Nidal Abd Aboelkher, Abdullah Saleh Bqies), ophthalmic assistants and nurses (Shahwan Ahmed Al zhrani, Eedah al-daadi, Ahmed Al-holaifi, Bashyer Abdullah, Faredah Dgrira Almuwalad), retinal photographers (Majed Nasser Al-Muhajeb, Jawaher Mohammad Al-Shahri), survey coordinators (Saleh Mohammed Alghamdi, Ziad Abdullbasit Mahboob) and data entry staff (Nino Papa Fucio).

    References

    1. ↵
      1. Pascolini D,
      2. Mariotti SP
      . Global estimates of visual impairment: 2010. Br J Ophthalmol. 2012;96:614–8.
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Tabbara KF,
      2. Ross-Degnan D
      . Blindness in Saudi Arabia. JAMA 1986;255:3378–84.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      Directorate of statistics MoH, Kingdom of Saudi Araboa. Chapter 1: A Review Of The Health Situation. Health Statistical Year Book 2006. Riyadh, Kingdom of Saudi Arabia: Darul Hilal, 2007:98.
    4. ↵
      International Diabetes Federation. IDF Diabetes Atlas. 4th edn. Brussels, Belgium: International diabetes Federation, 2009.
    5. ↵
      1. Al-Nozha MM,
      2. Al-Maatouq MA,
      3. Al-Mazrou YY,
      4. et al
      . Diabetes mellitus in Saudi Arabia. Saudi Med J 2004;25:1603–10.
      OpenUrlPubMedWeb of Science
    6. ↵
      1. Resnikoff S,
      2. Pascolini D,
      3. Mariotti SP,
      4. et al
      . Global magnitude of visual impairment caused by uncorrected refractive errors in 2004. Bull World Health Organ 2008;86:63–70.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Kuper H,
      2. Polack S,
      3. Limburg H
      . Rapid assessment of avoidable blindness. Community Eye Health 2006;19:68–9.
      OpenUrlPubMed
    8. ↵
      1. Polack S,
      2. Yorston D,
      3. Lopez-Ramos A,
      4. et al
      . Rapid assessment of avoidable blindness and diabetic retinopathy in Chiapas, Mexico. Ophthalmology 2012;119:1033–40.
      OpenUrlCrossRefPubMed
    9. ↵
      1. Chiang F,
      2. Kuper H,
      3. Lindfield R,
      4. et al
      . Rapid assessment of avoidable blindness in the Occupied Palestinian territories. PLoS One 2010;5:e11854.
      OpenUrlCrossRefPubMed
    10. ↵
      World Health Organization. Diabetes Mellitus. Report of a WHO Study Group Technical Report Series 727. Geneva: World Health Organization, 1985.
    11. ↵
      Scottish Diabetic Retinopathy Grading Scheme. 2007. v1.1. http://www.ndrs.scot.nhs.uk/ClinGrp/Docs/Grading%20Scheme%202007%20v1.1.pdf
    12. ↵
      1. al Faran MF,
      2. al-Rajhi AA,
      3. al-Omar OM,
      4. et al
      . Prevalence and causes of visual impairment and blindness in the south western region of Saudi Arabia. Int Ophthalmol 1993;17:161–5.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Rajavi Z,
      2. Katibeh M,
      3. Ziaei H,
      4. et al
      . Rapid assessment of avoidable blindness in Iran. Ophthalmology 2011;118:1812–18.
      OpenUrlCrossRefPubMed
    14. ↵
      1. Al Gamra H,
      2. Al Mansouri F,
      3. Khandekar R,
      4. et al
      . Prevalence and causes of blindness, low vision and status of cataract in 50 years and older citizen of Qatar-a community based survey. Ophthalmic Epidemiol 2010;17:292–300.
      OpenUrlCrossRefPubMed
    15. ↵
      1. Faal H,
      2. Minassian D,
      3. Sowa S,
      4. et al
      . National survey of blindness and low vision in The Gambia: results. Br J Ophthalmol 1989;73:82–7.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Dandona R,
      2. Dandona L
      . Socioeconomic status and blindness. Br J Ophthalmol 2001;85:1484–8.
      OpenUrlFREE Full Text
    17. ↵
      1. Abou-Gareeb I,
      2. Lewallen S,
      3. Bassett K,
      4. et al
      . Gender and blindness: a meta-analysis of population-based prevalence surveys. Ophthalmic Epidemiol 2001;8:39–56.
      OpenUrlCrossRefPubMed
    18. ↵
      World Health Organization. Informal Consultation on Analysis of Blindness Prevention Outcomes. Geneva: World Health Organization, 1998. WHO/PBL/98.86.
    19. ↵
      1. Yorston D,
      2. Gichuhi S,
      3. Wood M,
      4. et al
      . Does prospective monitoring improve cataract surgery outcomes in Africa? Br J Ophthalmol 2002;86:543–7.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      Photocoagulation treatment of proliferative diabetic retinopathy. Clinical application of diabetic retinopathy study (DRS) findings, DRS report number 8. The diabetic retinopathy study Research group. Ophthalmology 1981;88:583–600.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Turner R,
      2. Cull C,
      3. Holman R
      . United Kingdom prospective diabetes study 17: a 9-year update of a randomized, controlled trial on the effect of improved metabolic control on complications in non-insulin-dependent diabetes mellitus. Ann Intern Med 1996;124:136–45.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      Diabetes Control and Complications Trial Research Group. Progression of retinopathy with intensive versus conventional treatment in the diabetes control and complications trial. Ophthalmology 1995;102:647–61.
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Moss SE,
      2. Klein R,
      3. Kessler SD,
      4. et al
      . Comparison between ophthalmoscopy and fundus photography in determining severity of diabetic retinopathy. Ophthalmology 1985;92:62–7.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Saari JM,
      2. Summanen P,
      3. Kivela T,
      4. et al
      . Sensitivity and specificity of digital retinal images in grading diabetic retinopathy. Acta Ophthalmol Scand 2004;82:126–30.
      OpenUrlCrossRefPubMed
    25. ↵
      1. Williams GA,
      2. Scott IU,
      3. Haller JA,
      4. et al
      . Single-field fundus photography for diabetic retinopathy screening: a report by the American Academy of Ophthalmology. Ophthalmology 2004;111:1055–62.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Olson JA,
      2. Strachan FM,
      3. Hipwell JH,
      4. et al
      . A comparative evaluation of digital imaging, retinal photography and optometrist examination in screening for diabetic retinopathy. Diabet Med 2003;20:528–34.
      OpenUrlCrossRefPubMedWeb of Science
    View Abstract

    Footnotes

    • Funding This study was supported with funding from Fight for Sight (grant number 1765) and International Agency for the Prevention of Blindness, Eastern Mediterranean region (IAPB, EMRO).

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval LSHTM.

    • Provenance and peer review Not commissioned; externally peer reviewed.