Screening practices for diabetic retinopathy

Abstract

Diabetic retinopathy (DR) is a major cause of acquired blindness but is asymptomatic in its early stages. Therefore, there is an urgent need to establish a nation-wide efficient screening strategy for early detection, accurate diagnosis and timely treatment of vision threatening DR.

Keywords:

• diabetes
• diabetic retinopathy
• ophthalmoscopy
• screening
• telemedicine

Diabetic retinopathy (DR) ranks sixth among the causes of blindness in India [1]. The global prevalence of diabetes mellitus is estimated to increase by 54% by 2030 [2]. In India, the rate of increase is estimated to be 150% and so will DR [3]. The overall prevalence of DR is 28% in urban Chennai, India and 18% among those with diabetes [4]. In India, 80% of the population lives in rural areas. There is one ophthalmologist per 10,000 people and 70% of the ophthalmologist practices are in urban areas, with a very few trained in the diagnosis and treatment of DR [4]. Several screening programs are in place worldwide.

Recommendations

When to screen

The American Diabetes Association and the American Academy of Ophthalmology [5,6] recommend that adults and children >10 years of age with type 1 diabetes should have an initial dilated examination by an ophthalmologist within 5 years of the onset of diabetes. People with type 2 diabetes should have an initial dilated examination by an ophthalmologist at the time of their diabetes diagnosis, with yearly follow-ups or more frequently if retinopathy is progressing. Pregnant women with diabetes should have a dilated eye examination in the first trimester of pregnancy.

Who should screen

Ophthalmologist-based model

Conventionally, ophthalmologists have been performing screening for DR; this method comprises examining the entire retina by dilating the pupil, followed by indirect ophthalmoscopy. This method proves to be an advantage in situations where access to healthcare facility is feasible. However, the increasing diabetes epidemic will soon outrun the number of eye care providers available. Currently, several communities in India have limited or no access to ophthalmologic care.

Ophthalmologist-led model

This model primarily involves an optometrist or a trained technician to perform the preliminary examination at a site away from the base hospital such as in telescreening. Herein, digital photographs of the retina are acquired near patients’ home and are electronically transferred to a remote reading center where they can be evaluated by an ophthalmologist. It allows clinicians to detect DR in a non-ophthalmological setting and thus allows them to serve patients in rural, remote and hard-to-reach locations.

In either model, ophthalmologists play a significant role in communicating, both ophthalmologic findings and degree of retinopathy to the physician. The importance of tight control of blood glucose, blood pressure and blood lipid needs to be emphasized.

Existing screening models

The American Diabetes Association recommends that initial screening may be done by an ophthalmologist or a trained optometrist who is experienced in diagnosing DR. When there is a shortage of specialists, retinal photographs taken by trained photographer and/or reader may be used for follow-ups.

Telemedicine

Telemedicine has shown great promise as a public health intervention in DR screening. The American Telemedicine Association has published guidelines outlining essential components required for telescreening of diabetes that includes robust technology platform, quality assurance procedures, efficient workflows and compliance with legal and regulatory requirements [7].

Retinal imaging devices must meet current recommendations of image resolution (20 pixels per degree with a resolution of at least 10–15 μm) and must be validated.

How to screen

Various methods of detecting DR include direct and indirect ophthalmoscopy, stereoscopic color film fundus photography and mydriatic or non-mydriatic digital color photography.

Direct versus indirect ophthalmoscopy

The role of direct ophthalmoscopy in a screening has been evaluated in several studies. With an undilated direct ophthalmoscopy, the sensitivity has been reported to be less than 80% for diabetologists and non-ophthalmologists and it varied with the type of retinopathy lesions [8]. A lower sensitivity makes it less than ideal technology for screening. Indirect ophthalmoscopy is a promising technique and is a successful screening strategy; however, it requires trained individuals. In addition, images from both techniques cannot be saved and therefore, a reference is unavailable during follow-ups.

Mydriatic versus non-mydriatic fundus camera

Undilated 45° fundus photography for DR screening uses infrared system and therefore does not require pupil dilation [9]. The images are available for instant print and could be utilized in patients’ education, staff training and for auditing purposes. However, the quality of photographs is poor and the sensitivity is low for detecting DR in the periphery, thus making it less effective in detecting retinopathy lesions in the periphery. Traditional mydriatic fundus photography involves capturing images on 35 mm transparencies or films. Modern day cameras capture digital images of the fundus that aid in easy storage and transfer of images and are reported to have better sensitivity in identifying DR lesions, especially with mydriasis and is comparable to the traditional ophthalmoscopy [10]. The gold standard for DR detection has been the Early Treatment Diabetic Retinopathy Study (ETDRS) that uses 30°, seven standard field stereoscopic 35 mm color slides (7F-ETDRS) [11]. However, the 7F-ETDRS method can be expensive and time consuming, and makes it difficult to screen in rural areas [12]. The Liverpool Declaration, 2005, reported one-field photography to be adequate for photographic screening [13].

Novel screening techniques

Smart phones

The speedy evolution of smart phone technology and their ever-improving built-in camera technology has raised much interest in their use for medical and ophthalmic imaging. Various prototypes have been created to optically match the smart phone’s camera to a slit-lamp ophthalmoscope or to use it in conjunction with a condensing lens based on the principle of indirect ophthalmoscope [14]. The Bluetooth and Wi-Fi connectivity readily available on contemporary smart phones could allow seamless transmission of images. However, the field of view is limited when using mobile phones as a direct ophthalmoscope, and also the algorithms need to be validated on a robust population-based data.

Automated imaging

Although telescreening has greatly optimized the use of available manpower and other resources, it is still a daunting task for a designated ophthalmologist to screen thousands of images on a daily basis. Thus, the incorporation of automated screening software that would serve as a primary sieve to filter out the normal images is desirable. The automated system has high sensitivity and diagnostic accuracy that is comparable to that of specialists and expert readers. Automated methods result in approximately the same number of correct screening outcomes as manual methods, but at lower cost. But they need further validation on ethnically diverse populations [15].

A screening model for Indian population

Sankara Nethralaya, the premier eye institute of India in collaboration with Lions Club International Foundation and RD Tata Trust, Mumbai, India, launched a major DR screening program in South India [16]. The program aimed to create awareness among South Indians, conduct screening camps for diabetes and DR and to bring patients with sight-threatening retinopathy to the hospital for further investigations and management. The team comprises retinal specialists, genetic researcher, optometrists, research fellows, fundus photographer and social workers.

DR screening

Persons underwent dilated fundus evaluation with binocular indirect ophthalmoscope and +20D lens. Patients with narrow angles and sight-threatening DR were re-examined at the base hospital for further evaluation or treatment. People with non-sight-threatening DR and those with no DR were advised regarding follow-up.

We adopted this telescreening model to reach out to the rural populations who don’t have access to eye care facilities. With the exception of fundus evaluation, all other aspects of eye examination were performed by the paramedical staff that included optometrists. Our team had four ophthalmologists who performed dilated fundus evaluation using indirect ophthalmoscopy, to facilitate grading of DR.

We previously evaluated single-field 45° digital fundus photography in a telescreening model for DR [17]. We observed that with pupillary dilation, the gradability of the images improved with single-field photography. Single-field 45° photography cannot replace dilated retinal exam with indirect ophthalmoscopy; nevertheless, when combined with pupillary dilation, it can be utilized for telescreening of large rural populations. About 80% of the population lives in rural areas and about 70% of ophthalmologists practices are in urban areas. Therefore, we propose that a telescreening with an ophthalmologist-led model is preferable in India, to ensure adequate coverage of rural areas.

Barriers to screening

Several barriers such as low doctor to patient ratio, low patient education, lack of infrastructure to support services, lack of national policies present challenges to screening programs. Telemedicine approach is optimal in these settings. However, retinal imaging is plagued by high instrumentation costs. Retinal cameras continue to approach US$15,000–US$20,000. A true low-cost (less than US$1000) validated non-mydriatic retinal camera is still not available [18]. In southern India, a rural mobile telemedicine screening program has been established which aims to overcome some of the aforementioned barriers and provide examination services for populations who may not otherwise have access to such a program [19].

Cost–effectiveness of screening for DR

A teleophthalmology program conducted in a rural Indian community was cost-effective (US$1320 per quality-adjusted life-year) compared with no screening. Screening at a frequency of every 2 years was cost-effective, but annual screening was not. Screening can identify those at risk of visual loss and the cost involved is several times lower compared with disability benefits provided to those with visual impairment and blindness [20].

In conclusion, there is a need to create a diabetic network involving physicians, diabetologists, government and private hospitals and ophthalmologists. Integration of new technology with involvement of caregivers and diabetics will hold the key to success in achieving screening goals.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending or royalties.

No writing assistance was utilized in the production of this manuscript.