Article Text

Download PDFPDF

Availability and variability in guidelines on diabetic retinopathy screening in Asian countries
  1. Louis Zizhao Wang1,2,
  2. Carol Y Cheung1,3,
  3. Robyn J Tapp1,4,5,
  4. Haslina Hamzah1,
  5. Gavin Tan1,
  6. Daniel Ting1,
  7. Ecosse Lamoureux1,6,
  8. Tien Yin Wong1,6
  1. 1Singapore Eye Research Institute, Singapore National Eye Centre, Singapore, Singapore
  2. 2Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore
  3. 3Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong, Hong Kong, Hong Kong
  4. 4Department of Optometry and Vision Sciences, The University of Melbourne, Melbourne, Victoria, Australia
  5. 5School of Clinical and Applied Sciences, Leeds Beckett University, Leeds, UK
  6. 6Duke-NUS Medical School, National University of Singapore, Singapore, Singapore
  1. Correspondence to Professor Tien Y Wong, Singapore National Eye Centre, 11 Third Hospital Avenue, Singapore 168751, Singapore; wong.tien.yin{at}


Background Diabetic retinopathy (DR) is a blinding yet treatable complication of diabetes. DR screening is highly cost-effective at reducing blindness. Amidst the rapidly growing diabetic population in Asia, the prevalence of DR in the region is relatively less well known.

Aims To review existing national DR screening guidelines of 50 countries in Asia, compare them against the International Council of Ophthalmology (ICO) guideline, and summarise the prevalence rates of DR and sight-threatening DR (STDR) in these countries.

Methods We systematically searched for published guidelines from the National Guideline Clearinghouse and other databases, and contacted local diabetic and ophthalmological associations of all 50 Asian countries.

Results Eleven Asian countries have published relevant guidelines, nine of which pertain to general diabetes care and two are DR-specific, covering less than half of Asia's population. The median DR prevalence among patients with diabetes is 30.5% (IQR: 23.2%–36.8%), similar to the USA and the UK. However, rates of STDR are consistently higher. All guidelines from the 11 Asian countries fulfil the ICO standard on when to start and repeat screening, except for screening interval for pregnant patients. However, only 2 of the 11 guidelines fulfil the ICO referral criteria and 6 partially fulfil. A third of the recommendations on screening process, equipment and personnel is either unavailable or incomplete.

Conclusions Countries in Asia need to establish more comprehensive and evidence-based DR screening guidelines to facilitate the execution of robust screening programmes that could help reduce DR-related blindness, improve patient outcomes and reduce healthcare costs.

  • Retina
  • Epidemiology

Statistics from

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.


Diabetes mellitus is one of the world's major epidemics affecting 415 million people and is expected to expand to 642 million by year 2040.1 Diabetic retinopathy (DR), a potentially blinding complication of diabetes, is the leading cause of preventable blindness in the adult population.2 In patients with diabetes, the global prevalence of any DR is 34.6% and 1 in 10 have sight-threatening DR (STDR), which commonly includes severe non-proliferative (NPDR), proliferative DR (PDR) and diabetic macular oedema (DMO).3 Once diagnosed with DR, 11% progress to the sight-threatening stage every year, making DR a significant public health challenge.4

The natural history of DR is asymptomatic initially, but can progress to total blindness if not promptly treated.5 Due to the rising prevalence of DR and the clinical latency early in its course, it is important to conduct timely screening so that appropriate treatment can be implemented to prevent blindness. Primary interventions through glycaemic and blood pressure control and secondary interventions via laser treatment or intravitreal injections can reduce the incidence and slow down the progression of DR.3 ,6 ,7

Asia is home to 60% of people with diabetes in the world.8 Of the 10 countries with the largest diabetic population worldwide, half are in Asia—namely China, India, Indonesia, Japan and Bangladesh. In China, it is estimated that 12% of adults have diabetes and half are prediabetic.9 Currently, national data on DR in Asian countries are limited due to the lack of large population-based studies and differences in study methodologies.10 However, available reports in the region show that its prevalence rates are as high as 43.1%.11 In particular, high rates of undiagnosed diabetes also contribute to the overall impact of DR in Asia, as undiagnosed patients could be more predisposed to visual complications due to poorer care. Reported prevalence of these patients in low income, middle income Asian countries account for over half of the total diabetes population,12–17 giving rise to a hidden disease burden that would continue to proliferate with the lack of timely intervention and risk factor control.

Given the weight of diabetes and DR in Asia, there is a pressing need for national DR screening guidelines to promote retinal evaluation in all patients with diabetes, in order to reduce the rates of DR-related blindness. Studies have consistently demonstrated significant improvements in visual outcomes with national screening guidelines and programmes in place.18–21 However, it is unclear how available such guidelines are in Asia and whether available recommendations follow evidenced-based practices.

In this review, we determined the availability and variability of national DR screening guidelines in Asian countries, summarised their recommendations and compared them with the gold standard, Guidelines For Diabetic Eye Care by the International Council of Ophthalmology (ICO), which is an authoritative international organisation consisting of 120 national and subspecialty member societies from around the globe.22 Areas for improvement have been identified and may serve as a feedback for future versions of guidelines, in an effort to effectively combat DR-related vision loss in the region.


We performed a systematic review of existing literature to determine the availability of guidelines on DR screening in Asia, which per United Nations definition, comprises 50 member countries and states.23 Search for existing guidelines on diabetes and DR care in each country was performed by using the PubMed database, the National Guideline Clearinghouse and the Google search engine on keywords: ‘country of interest’ AND (‘diabetes’ OR ‘diabetic retinopathy’ OR ‘screening’ OR ‘guidelines’ OR ‘clinical practice guidelines’). The official websites of the diabetes associations and ophthalmology societies of each country, where available, were also explored in search of the relevant guidelines. After which, three rounds of email requests were directed to the respective national diabetic and ophthalmological authorities to add to the pool of identified guidelines.

Only guidelines published by national health authorities that address the entire population rather than being specific to only part of a country, a community or a research study group, were included. Guidelines of any language were accepted and translated.

Based on the guidelines obtained, information on diabetes and DR disease burden, recommended time to start screening, frequency of re-evaluation, referral criteria and the screening process, methods and personnel was summarised.


A total of 11 national guidelines fulfil the inclusion criteria, of which 10 were obtained from initial literature review and 1 via email correspondence. Of these, nine were published on general diabetes care while two were specific to the management of DR (table 1). The median prevalence of DR among patients with diabetes is 30.5% (IQR: 23.2–36.8%), similar to the USA and the UK. However, the rates of STDR in these Asian countries are consistently higher (5.2% to at least 20%i), compared with 4.4% in the USA and 3.4% in the UK (figure 1).

Table 1

Overview of diabetes guidelines in Asia

Figure 1

Prevalence of diabetic retinopathy (DR) and sight-threatening diabetic retinopathy (STDR) in Asia. PDR, proliferative diabetic retinopathy.

When to start screening

All identified guidelines share the consensus of starting DR screening for patients with type 2 diabetes (T2DM) at diagnosis. For those with type 1 diabetes (T1DM), six guidelines recommend screening to be commenced within 5 years of diagnosis, while three suggest during the peripubertal period. For pregnant patients with pre-existing diabetes, three guidelines recommend starting screening in the first trimester only, three suggest prior to conception only, while three accept both options.

Frequency of screening

Rather than the type of diabetes, the frequency of follow-up screening is determined based on the severity of DR, which is an important criterion influencing referral decisions (table 2). The ICO makes recommendations using the International Clinical Diabetic Retinopathy and Diabetic Macular Oedema Disease Severity Scale (referred hereafter as the International Classification)49 based on countries' resource availability, that is, low, intermediate or high. Separate recommendations are available for pregnant patients with diabetes, regardless of resource settings. Five guidelines adopt the International Classification like the ICO, two follow the UK National Screening Committee (NSC) Classification,50 while the remaining four guidelines have not specified any classification system used. Ten guidelines fulfil the ICO standard, while one did not provide any screening frequency recommendation for DR of all severities. For no DR, the ICO recommends repeating screening every 1–2 years regardless of resource settings. All 10 guidelines recommend annual screening for patients with no DR, half of which also accept a longer screening interval of ≥2 years, providing more flexibility for countries with low and intermediate resource settings. For mild NPDR, the ICO suggests repeating screening every 1–2 years for low/intermediate-resource settings and 6–12 months for high-resource settings. Ten guidelines fulfil this standard by recommending screening annually or more frequently, especially for patients at high risk of DR progression.

Table 2

Recommended frequency of screening and referring criteria

The ICO recommends referring patients with moderate NPDR. Apart from the two guidelines fulfilling the international standard of referring from the preproliferative stage on the NSC Classification,50 only two guidelines meet the ICO standard. Excluding the guideline without recommendation, the other six guidelines recommend screening annually or more, but only to start referring after patients have already progressed to the sight-threatening stage.

Pregnant patients with pre-existing diabetes

Depending on results of the initial examination, the ICO recommends repeating retinal assessment at 28 weeks for pregnant patients with no DR, and at 16–20 weeks for those with DR. This is fulfilled by three guidelines and partially fulfilled by five for providing vague recommendations without being specific, while the remaining three guidelines have not provided any recommendation. For guidelines that have partially fulfilled the ICO standard, three advise to screen at physician's discretion and two suggest ‘close follow-up’, all without further elaboration.

Criteria for referral

The ICO guideline recommends three key referral criteria: (1) visual acuity below 6/12 (20/40) or having any symptomatic vision complaints; (2) from moderate NPDR onwards based on the International Classification; (3) if visual acuity or retinal examination cannot be obtained at the screening examination. Two guidelines meet the ICO standard by stating all three referral criteria, six partially achieve the standard by stating only one or two criteria, while three do not meet the ICO standard.

A high quality referral system is pivotal in ensuring the accurate selection of all at-risk individuals who are in need of specialised care for urgent interventions. In order to adhere to the three key referral criteria as recommended by the ICO, the prerequisite is to establish a screening programme with a well defined and comprehensive screening process, supported by the appropriate equipment and personnel.

Screening process, equipment and personnel

As outlined by the ICO guideline, a typical screening session should include a history taking and a physical examination component. Six fulfil the ICO standard for having both components and three partially fulfil by including either component. The ICO recommends using non-mydriatic retinal photography (RP) as the core screening equipment and accepts other modalities such as direct/indirect ophthalmoscopy. Six guidelines choose RP as the first-line screening modality and one recommends doing indirect ophthalmoscopy. For screening personnel, all nine guidelines with specified recommendations fulfil the ICO standard, of which five require only trained personnel, three recommend ophthalmologist or optometrist, and one accepts either (table 3).

Table 3

Screening process, equipment, personnel

In general, the 11 Asian countries have good interguideline agreement and are in line with international standards on the time to commence screening and frequency of follow-up evaluations, except for the follow-up screening intervals for pregnant patients. Only two guidelines fully fulfil the ICO standard for criteria to refer, and a third of the recommendations for screening process, equipment and personnel are either unavailable or incomplete (table 4).

Table 4

Comparison of guidelines from 11 Asian countries with the International Council of Ophthalmology Guideline


Availability and variability of screening guidelines

Guidelines in Asia are lacking in availability with less than a quarter of Asian countries having relevant available guidelines. Furthermore, only two national guidelines are written specifically on the care of DR. A comprehensive guideline on DR screening should include recommendations on when to conduct the first and subsequent screenings, the criteria for referral and the screening process, personnel and equipment used. While a comprehensive guideline does not necessarily equate to having a robust screening programme, it serves as the cornerstone for establishing a well structured and cost-efficient nationwide screening programme. Without having national guidelines specific to DR screening, countries could become more vulnerable to DR-related blindness. Ideally, screening programmes should be tailored to each country based on local epidemiological and socioeconomic contexts. Nevertheless, before the emergence of local evidence to support such initiatives, countries should adapt from established guidelines by authoritative bodies such as the ICO, which caters to all countries with varying resource levels. Ultimately, the goal is to strive towards early detection and prompt intervention to achieve a reduction in DR-related blindness in the most cost-efficient way.

The number of patients with diabetes and DR in Asia is increasing at an incredible speed, especially in low income, middle income countries such as China and India. For instance, China currently has 10.6% of its population with diagnosed diabetes. This number has more than doubled from 4.5% in the past 6 years, and is catching up with developed counterparts like the USA, where the diabetes prevalence rate has remained stable at 12% over the same period.1 ,52 In the same way as for diabetes, DR used to be considered as relatively infrequent in low income, middle income countries. However, the rise in regional diabetes prevalence and the rapid rates of urbanisation, together with a possible genetic predisposition for diabetes in certain Asian ethnicities, are leading to similar if not higher rates of DR in Asia compared with the West.13 ,53 ,54 The result of a more severe problem of STDR in Asia (figure 1) could be a reflection of the lack of routine retinal evaluation and timely treatment of DR in early diabetes in the region, resulting in its late diagnosis only at the symptomatic stage.

Identified guidelines unanimously agree to start screening at diagnosis for patients with T2DM, and are dichotomised based on differing evidence for patients with T1DM (peripubertal vs within 5 years of diagnosis) and pregnant patients with pre-existing diabetes (first trimester vs preconception). The diagnosis of T1DM is often made following the acute presentation of symptoms at disease onset, while the onset of T2DM could have been present for years before diagnosis is made.55 Approximately 20% of patients with newly diagnosed T2DM have some evidence of DR, justifying the need to conduct screening at or shortly after the diagnosis of T2DM.56 ,57 Given that early detection is possible in T1DM, it is not necessary to start screening immediately after diagnosis. Evidence from the Wisconsin Epidemiologic Study of Diabetic Retinopathy showed a surge in incidence of DR following 5 years into the course of T1DM, supporting DR screening to be commenced within 5 years of diagnosis.58 Alternatively, screening during or shortly after puberty is also acceptable as it is uncommon for patients with T1DM below the age of 12 years to develop DR.59 ,60 Pregnancy can worsen the severity of pre-existing diabetes which can bring about an earlier onset of DR.61 Up to 85% of pregnant patients can experience progression of DR during pregnancy,62 ,63 with the highest risk of DR exacerbation in the second trimester,63 justifying an initial screening be done in the first trimester. Achieving an optimal glycaemic and blood pressure control during pregnancy may help reduce the risk of long-term DR progression, hence starting screening before conception can also be desirable if adequate resources are available.62 ,64 However, a transient paradoxical worsening of DR could be observed following the initiation of intensive insulin therapy, as explained by the ‘early worsening phenomenon’.65 ,66 This is especially so in women with poorly controlled diabetes at baseline, who may become increasingly motivated to optimise their diabetes control during the pregnancy, thus putting themselves at higher risk of early worsening. Therefore, close ophthalmological assessment before and during pregnancy is warranted in women with diabetes.

Regarding screening intervals for patients with no DR, all identified guidelines consistently recommend repeating screening annually. Although this is accepted by the ICO, internationally, there is a recent move towards biennial screening given the little difference in outcomes shown compared with screening annually.4 ,67–69 The ICO currently also supports repeating screening up to 2 years. Along with the lead of key international guidelines,22 ,70 it is likely that more countries will start adopting a biennial screening approach for patients with no DR in the future.

Significant differences have also been observed in the threshold for referral based on disease severity. Out of the five guidelines that adopt the International Classification, four suggest referring patients from severe NPDR onwards, but the ICO recommends an even earlier referral at the moderate stage, before the onset of STDR occurs. Only four guidelines fulfil this criterion, highlighting the need for policy makers to increase the stringency of their referral criteria based on the latest ICO standard. Referring patients at the moderate NPDR stage would allow for timely implementation of treatment for patients who are at risk of developing STDR, before the actual decline in vision sets in.

In addition, guidelines are generally vague on what a session should entail and how to conduct such sessions. Particularly, three guidelines suggest ‘comprehensive evaluation’ for screening without further elaboration. This could potentially create confusion among screening personnel and compromise on the quality of screening unnecessarily. However, this can be prevented if guidelines include a more detailed section on the proper screening process, by making reference to the ICO guideline, such that a standardised screening format can be established that allows screening personnel to easily follow.

The ICO endorses a wide variety of screening equipment including ophthalmoscopy, RP and slit-lamp. However, non-mydriatic RP is the preferred method that is only recommended by one guideline, while six mentioned RP without stating whether to perform mydriasis. Compared with RP with mydriasis, non-mydriatic RP is safer and faster by doing away with pharmacological pupil dilation. Being highly sensitive and specific for detecting both early and referable DR, it has been shown to increase access to screening and proven to be cost-effective in community screening programmes.71–76 Recently, RP has also been increasingly applied in the field of teleophthalmology, especially in areas with limited resources and poor access to eye care, achieving significant reductions in DR-related visual impairment.18 ,77 ,78 In the event that RP is unavailable or images are unreadable, ophthalmoscopy can be used as an alternative for screening.79 ,80

Finally, nine guidelines recommend well trained personnel, optometrist or ophthalmologist to conduct fundal examination, meeting the ICO standard. DR screening has usually been performed by eye specialists, due to concerns for poor detection rates by non-ophthalmologists.81 However, recent evidence has established a role for screening by non-ophthalmologists, given prior training in grading DR severities.82–87 The ICO has endorsed an online self-educational programme on assessing DR severities, currently available in six languages, serving as a training platform and a guide for countries intending to set up similar courses.88

Barriers to effective DR screening

A successful screening programme should entail its smooth execution by healthcare providers and active participation from the public. Resource constraints may result in the shortage of high-quality screening guidelines and hence limit on the provision of screening services. However, even some of Asia's wealthiest countries including Qatar, Macau, Brunei and Kuwait have no guidelines available. Furthermore, reports of severe underutilisations of DR screening services show an uptake rate of less than 20%.89 Factors impeding screening uptake include the lack of awareness on DR, transport safety concerns and language barriers for patients, as well as the lack of medical advice from eye care providers.90–95


DR is potentially blinding and is rapidly growing in Asia, where 60% of the world's population resides. Despite facing similar rates of DR and even higher rates of STDR than the West, fewer than a quarter of Asian countries have available DR screening guidelines, serving less than half of Asia's population. Existing guidelines need to be evidence-based and more comprehensive to provide a basis for screening programmes to be established. We urge policy makers to create new guidelines and refine existing ones so that populations at risk may benefit from timely intervention and avoid DR-related blindness. The ICO guideline serves as a benchmark defining the standards of care for countries with different resource settings, thereby allowing targeted cost-effective screening strategies to be devised.


The authors thank the following people for providing valuable information on the availability of DR screening guidelines in the respective countries: Professor Ningli Wang, Dr Fenghua Wang, Dr Yiquan Yang (Beijing Tongren Hospital); Dr Petros Aristodemou (Cyprus Ophthalmological Society); Professor Dexter Leung, Dr Pui Pui Yip (Hong Kong Ophthalmological Society); Professor Hamid Ahmadieh (Iranian Society of Ophthalmology); Dr San Myint and Dr James Muecke (Vision Myanmar Program); Dr Lila Raj Puri (Nepal Ophthalmic Society); Professor Madhur Dev Bhattarai (Nepal Diabetes Association); Dr Rima Tan (Diabetes Philippines); Professor Maged A Abu-Ramadan (The Palestinian Ophthalmological Society); Professor Klaus G Parhofer in Turkmenistan (University of Munich).



  • Contributors LZW, CYC and TYW were involved in (1) conception and design, acquisition of data, or analysis and interpretation of data; (2) drafting the article or revising it critically for important intellectual content; and (3) final approval of the version to be published. RJT, HH, GT, DT and EL have contributed substantially to (1) conception and design, acquisition and interpretation of data; (2) revising it critically for important intellectual content.

  • Funding National Medical Research Council (NMRC) Singapore, grant number STaR/0016/2013.

  • Competing interests None declared.

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

  • i 20% represents PDR only.