C L I N I C A L

A N D

E X P E R I M E N T A L

OPTOMETRY cxo_746

311..327

REVIEW

Telemedicine and ocular health in diabetes mellitus

Clin Exp Optom 2012; 95: 311–327 Sven-Erik Bursell*† PhD Laima Brazionis§ PhD MHN BSc (Opt) Grad Dip Epi Biostats Alicia Jenkins* MD FRACP * The University of Melbourne, Department of Medicine, St Vincent’s Hospital, Melbourne, Australia † University of Hawaii, John A Burns School of Medicine, Telehealth Research Institute, Hawaii, USA § University of Adelaide, Discipline of Public Health, School of Population Health and Clinical Practice, Adelaide, Australia E-mail: [email protected]

Submitted: 23 February 2012 Revised: 26 March 2012 Accepted for publication: 30 March 2012

DOI:10.1111/j.1444-0938.2012.00746.x Teleretinal/teleophthalmological programs that use existing health information technology infrastructure solutions for people with diabetes increase access to and adherence to appropriate eye care. Teleophthalmological studies indicate that the single act of patients viewing their own retinal images improves self-management behaviour and clinical outcomes. In some settings this can be done at lower cost and with improved visual outcomes compared with standard eye care. Cost-effective and sustainable teleretinal surveillance for detection of diabetic retinopathy requires a combination of an inexpensive portable device for taking low light-level retinal images without the use of pharmacological dilation of the pupil and a computer-assisted methodology for rapidly detecting and diagnosing diabetic retinopathy. A more holistic telehealth-care paradigm augmented with the use of health information technology, medical devices, mobile phone and mobile health applications and software applications to improve health-care co-ordination, self-care management and education can significantly impact a broad range of health outcomes, including prevention of diabetes-associated visual loss. This approach will require a collaborative, transformational, patient-centred health-care program that integrates data from medical record systems with remote monitoring of data and a longitudinal health record. This includes data associated with social media applications and personal mobile health technology and should support continuous interactions between the patient, health-care team and the patient’s social environment. Taken together, this system will deliver contextually and temporally relevant decision support to patients to facilitate their well-being and to reduce the risk of diabetic complications.

Key words: diabetes, retinopathy, telehealth

The incidence and prevalence of both type 1 diabetes and type 2 diabetes is increasing.1 In 2007–2008 an estimated 898,800 Australians had diagnosed diabetes (87,100 with type 1 and 787,500 with type 2).2 In Aboriginal and Torres Strait Islander peoples, diabetes prevalence is over threefold higher than that in nonindigenous populations.2

Diabetic complications and related conditions substantially increase the economic, social and personal burden of diabetes,3,4 which is the most common cause of blindness in working-age adults, predominantly due to diabetic retinopathy. Currently available medical therapies, such as glucose, blood pressure and lipid control, and non-smoking can retard

© 2012 The Authors Clinical and Experimental Optometry © 2012 Optometrists Association Australia

and sometimes reverse diabetic retinal damage,5–7 and for most people with latestage diabetic retinopathy laser therapy8 and intraocular anti-vascular endothelial growth factor drugs9 can reduce further visual loss. Unfortunately, diabetic retinopathy is often asymptomatic in its most treatable stages and many people with diabetes do Clinical and Experimental Optometry 95.3 May 2012

311

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

not receive regular eye care. Even in affluent countries there are not enough eyecare specialists to provide in-person care and education to each person with diabetes. The cost and time required to train and equip such clinicians is large and their geographic maldistribution (with a predominance located in major cities) is an additional challenge. Telemedicine might be part of the solution. The American Telemedicine Association defines telemedicine as ‘the use of medical information exchanged from one site to another via electronic communications to improve patients’ health status’ (http://www.americantelemed.org/i4a/ pages/index.cfm?pageid=3333). Closely associated with telemedicine is the term ‘telehealth,’ which is often used to encompass a broader definition of remote health care that does not always involve clinical services. Videoconferencing, transmission of still images, e-health including patient portals, remote monitoring of vital signs, continuing medical education and nursing call centres are all considered part of telemedicine and telehealth. The use of telemedicine could facilitate eye screening, education, diabetes care and the efficient triage of patients to the appropriate practitioner. Given the disease burden of diabetes and related visual loss and given that the presence of diabetic retinopathy often signals the presence or increased risk of diabetic nephropathy, neuropathy, cardiovascular disease and premature mortality,10 diabetic retinopathy represents a potential model for optimal chronic disease management using telemedicine and telehealth. It is critical that telemedicine and telehealth programs match the quality of care expected in traditional clinical settings. Therefore, rigorous quality assurance and appropriate education and certification of the users and monitoring of quality are essential.

DIABETES AND EYE CARE Diabetic retinopathy and other common age-related causes of visual impairment, such as cataracts and glaucoma, which are Clinical and Experimental Optometry 95.3 May 2012

312

more prevalent in those with diabetes, can be detected by remote ocular imaging.11 Despite widespread recommendations,12,13 approximately 50 per cent of people with diabetes do not have regular eye examinations. McCarty and colleagues14,15 reported that nearly half the people with diabetes (69,000) in the major Australian city of Melbourne (population 3,471,625 in 2001) are not receiving appropriate eye care, despite universal health care and geographical proximity to eye specialists. Individual factors influencing non-adherence to eye care include type of health-care provider, sociodemographics, diabetes type and duration and diabetes education.16 People were more likely to be non-adherent to recommended eye assessments if they did not receive an eye-care recommendation from a medical practitioner (as opposed to another type of health-care professional), were younger, had type 2 diabetes and did not use insulin, had diabetes of short duration and/or did not have access to or participate in a formal diabetes program. In community health-care settings, affordability of goods and services and ability to pay are also barriers to appropriate care. Community-level factors affecting nonadherence include proximity to health care and cultural barriers.17 TELEOPHTHALMOLOGY/ TELERETINAL SURVEILLANCE Teleophthalmological systems are relatively mature telehealth applications with teleradiology being the most mature of the telehealth applications. Table 1 summarises telehealth applications at various stages of maturity. Several other telehealth applications might also facilitate diabetes care and impact ocular health by improvements in risk factor control, such as glucose, blood pressure, lipids, weight and smoking. Teleophthalmological systems can facilitate eye care while addressing access to care and cost issues. Such programs involve taking retinal images at the point of encounter, using non-mydriatic retinal imaging systems and trained retinal imagers. In addition, images of the exter-

nal eye can be taken and visual acuity can be tested using the retinal imager. The retinal images are then securely transmitted to centralised reading centres, where an assessment of ocular status is performed by trained and certified retinal graders. The reports from the reading centre are transmitted back to the originating site and management of the patient remains with the local general practitioner, who will develop a management plan. Thus, by transferring some routine eye examinations from eye-care specialists to other settings or non-specialists, teleophthalmology can improve accessibility of specialty eye care for follow up and treatment of highrisk patients with sight-threatening abnormalities. Such a system can reduce travel and its related costs and also reduce waiting times for specialised eye care. Costeffective and sustainable teleretinal surveillance to facilitate the management of diabetic retinopathy and other eye diseases ideally requires a combination of an affordable portable device for taking low lightlevel retinal images (a data collection device) without the use of pharmacological pupil dilation (so as to be minimally intrusive to the patient) and a computer-assisted methodology for rapidly detecting and diagnosing diabetic retinopathy (a data analysis application) and other sightthreatening ocular disorders, such as macular oedema, glaucoma, cataracts and trachoma. The past 15 years have seen the development and international implementation of a number of telemedicine-based surveillance systems for assessment of diabetic retinopathy.18–35 In the UK, teleretinal screening for diabetic retinopathy has been adopted as a program within the National Health Service.36 In Australia, retinal digital photography is widely used by optometrists and ophthalmologists. Teleretinal screening programs for diabetic retinopathy have been implemented but the service in Australia is not as widespread as in Europe or the US, primarily because the service is not reimbursed.37 Teleretinal surveillance systems represent a unique integration of telemedical technology and clinical programs designed to increase the access of patients © 2012 The Authors

Clinical and Experimental Optometry © 2012 Optometrists Association Australia

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

Modality

Comments

Medical call centres

Most mature with a well established business model providing online care virtual visits with health-care professionals: http:// www.3gdoctor.com/; http://wwc.americanwell.com/.

Peer to peer consultation and second opinion

Video teleconferencing, often difficult to manage logistically and cost-effectiveness is equivocal.

Teleradiology

Mature application with rigorous standards and workflow established, such as sharing of images with consultants. Current standard allows images to be read in urgent and emergent situations without a radiologist on site.

Teleophthalmology

Also a mature application primarily focused on diabetic retinopathy surveillance. There are well established rigorous standards in place.

Teledermatology

Primarily store and forward technology where the picture and related information is captured and sent to the consultant who reviews and responds. A hybrid model of store and forward and video teleconferencing shows best route for reimbursement.

Telepsychiatry

There are a number of successful programs using videoconferencing, particularly child psychiatry where resources and access are limited.

Telepharmacy

Remote delivery of pharmacy services using pharmacy technician or vending machine services coupled with videoconference between patient and pharmacist. Also inpatient pharmacy consultations covering pharmacy services in smaller hospitals, often at night.

Telepathology

Usually a store and forward application with the limitation that the viewing pathologist cannot remotely control the microscope to view a specimen as this requires high bandwidth. Ongoing concerns regarding quality of service.

Telesurgery and health-care assistive robots

Telesurgery is the use of telecommunications and information technology to perform surgical procedures on patients using virtual reality and telerobotics. It represents a nascent telehealth technology with promise. Health-care assistive robots are mobile robots that move and navigate to assist nurses and other care providers with routine, repetitive or physically strenuous tasks and also represents a promising telehealth technology.

Telestroke

Dependent on CT and MRI image transfer. Used to remotely triage patients for clot reducing drugs. Study results indicate that telestroke management is as effective as bedside assessment.21

Telerehabilitation

This is the delivery of rehabilitation services over telecommunication networks and the internet. The services fall into two categories: clinical assessment (the patient’s functional abilities in his or her environment) and clinical therapy. Telerehabilitation services include neuropsychology, speech–language pathology, audiology, occupational therapy and physical therapy. Telerehabilitation allows experts in rehabilitation to engage in remote clinical consultations.

Care management

Allows the introduction of the ‘virtual team’ into the management of chronic conditions. Represents an extension of the PCEHR and eHealth for disease management, care co-ordination and a more continuous communication between the patient and the care team.

Education

Distance learning traditionally using interactive videoconferencing but now moving to real-time webinars and a variety of interactive store and forward self-paced learning techniques. Primarily professional education but an increasing focus on delivery of patient-centric education including dynamic health coaching, lifestyle and wellness and use of text messaging.

Home and remote monitoring

The technology for remote monitoring is available (including weight, blood pressure, blood glucose, physical activity, motion and fall sensing, ECG monitoring, sleep quality monitoring, electronic pill box), but there is limited experience on how this technology will change the delivery system in managing chronic diseases. An electronic pill box for medication adherence and medication reminders resulting in reduced hospitalisations, costs and improved quality of life also reduced caregiver stress.

Personal health records (PCEHR)

The ‘personal health record’ is a concept that is not fully mature and is represented by Microsoft Health Vault. The PHR and its extension as the ‘patient centred medical home’ is growing in acceptance; however, and is undergoing demonstration testing in Australia.

Patient-centred medical home or virtual medical home

Represents an interactive approach to providing comprehensive primary care and chronic disease care management, while facilitating partnerships between individual patients and their personal providers and, when appropriate, the patient’s family. The provision of medical homes might allow better access to health care, increase satisfaction with care and improve health.

Mobile health (mHealth)

Mobile health is the practice of medicine and public health, supported by mobile devices. This involves remote monitoring that is not home based. The term is most commonly used in reference to using mobile communication devices, such as smart mobile phones and PDAs (iPad or other smart tablet devices) for health services and information: http://www.bodymedia.com/; http://www.medapps.com/; http://www.myglucohealth.net/.

Social support networks

In today’s mobile environment people are connected, so their health is connected. Health behaviours spread from person to person and health care can harness that phenomenon to transform the health and wellness of large populations (http:// news.theage.com.au/national/social-networking-may-improve-our-health-20080116-1mbe.html). Mobile social network groups have been effective in weight loss and smoking cessation and aging at home programs. For example, a group of home-bound patients can watch the same film and discuss over dinner with those connected through the virtual support group connectivity.

Simulation and games for health

Gaming is increasingly being used in all aspects of health, from increasing compliance with medications to physical and visual therapy. For example, the Nintendo Wii Fit (http://wiifit.com/) helps increase activity levels (especially seniors) through interactive fitness games. Simulation is also being used in medical professional education (Burn Centre simulation prepares medical community for mass casualties and burns as the result of a major catastrophe: https://www.burncentretraining.com/).

CT: computed tomography, ECG: electrocardiogram, MRI: magnetic resonance imaging, PCEHR: personally controlled electronic health record, PHR: personal health record

Table 1. Telehealth specialty areas

© 2012 The Authors Clinical and Experimental Optometry © 2012 Optometrists Association Australia

Clinical and Experimental Optometry 95.3 May 2012

313

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

Figure 1. Retinal fields obtained for the clinically validated JVN teleophthalmology system

with diabetes to eye-care programs and increase the efficiency of providing this care according to existing clinical guidelines.13 Using commercially available cameras designed for digital video image acquisition, non-mydriatic digital images of different 45 degree areas or fields of the retina, including the disc, macula, temporal vascular arcades and the nasal retina18 are taken (Figure 1), then transmitted electronically to centralised reading centres for evaluation.18 Unlike standard evaluations for diabetic retinopathy, the reading centre can be close to the image acquisition site or across the world. Furthermore, pharmacological dilation of the pupil need only be considered in rare circumstances, because image acquisition of the retina is facilitated by using the lowest light level possible for adequate exposure and retinal image quality. This imaging process improves patient comfort and shortens the time needed for acquiring retinal images. The less intrusive process Clinical and Experimental Optometry 95.3 May 2012

314

facilitates patient compliance to appropriate eye care. VALIDATION OF TELEOPHTHALMOLOGICAL SYSTEMS The credibility of a teleophthalmological program relies on validation of the scope of service that will be provided. Multi-centre, national clinical trials have provided evidence-based, clinical guidelines for diagnosis and management of diabetic retinopathy.38,39 Telehealth programs for diabetic retinopathy should clearly define program goals and performance in terms consistent with locally accepted clinical standards. The Early Treatment Diabetic Retinopathy Study (ETDRS) 30-degree, stereo-seven standard field, colour, 35 mm slides and the International Clinical Diabetic Retinopathy Disease Severity Scale38,39 provide the benchmark for evaluating diabetic retinopathy and macular

oedema. Telehealth programs for diabetic retinopathy should demonstrate an ability to compare favourably with ETDRS film or digital photography, as reflected in kappa values for agreement of diagnosis and sensitivity and specificity of diagnostic levels for diabetic retinopathy and macular oedema.18,40,41 If image quality is not sufficient to achieve a diagnosis of diabetic retinopathy, this is considered a positive finding for the presence of retinopathy in telehealth programs, requiring further assessment by eye-care specialists. Ungradable image rates are a ‘quality assurance’ measure and acceptable frequencies of ungradable cases must be established through protocols and policies propagated by the program or organisation.42 Furthermore, patients with unobtainable or unreadable images should be either promptly re-imaged or referred for evaluation by an eye-care specialist. One study reported that the majority of patients referred due to unreadable images actually had ocular disease that would have resulted in referral if adequate images had been obtained.43 Teleophthalmological protocols (nonmydriatic or mydriatic) should state the standards used for validation and the datasets used for comparison. Generally, a higher rate of unreadable photographs has been reported through undilated versus dilated pupils.44,45 People with diabetes often have smaller pupils and a greater incidence of cataracts, which might limit image quality if performed through an undilated pupil. Pupillary dilation is associated with a very small risk of angle-closure glaucoma. Programs using dilation need a clearly defined protocol to recognise and address this potential complication. As angle closure might occur hours after pupil dilation, the patient should be provided with phone contacts and instructed to call if symptoms occur, and the staff should be able to advise the patient appropriately. A recent publication ‘Telehealth practice recommendations for diabetic retinopathy’46 documented four categories of validation of telehealth programs for diabetic retinopathy. It recognised that each validation category would impact hard© 2012 The Authors

Clinical and Experimental Optometry © 2012 Optometrists Association Australia

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

ware and software technology, the level of staffing and support, the clinical workflow, the clinical outcomes, quality assurance and the business case. Category 1 validation allows differentiations between retinas with no or minimal diabetic retinopathy (ETDRS level 20 or less) and those with more than minimal diabetic retinopathy. Category 2 validation allows differentiation between patients who do not have sight-threatening retinopathy and those who potentially do (ETDRS level 53 or worse). Patients with sight-threatening retinopathy generally require prompt referral and treatment. Category 3 validation can identify a defined set of ETDRS levels of diabetic retinopathy and macular oedema that allows appropriate follow up and treatment plans. This level of validation allows patient management to match clinical recommendations based on dilated pupil clinical examinations. Category 4 validation indicates a system that matches or exceeds ETDRS protocols for diabetic retinopathy and indicates a system that can replace ETDRS photography in any clinical or research program. The validation of a teleophthalmological system does not obviate the provider’s responsibility for the care of the patient.47 Providers who perform or oversee teleretinal programs are responsible for patient care and safety, the quality of the retinal review process and the resultant management plan for that patient. Thus, regardless of a patient’s location and an image’s origin, providers should ensure the quality of medical images and policies that affect patient care and safety.48–50 Teleophthalmological systems have been validated through peer-reviewed publications for diabetic retinopathy, macular oedema and other pathologies such as risk for glaucoma.25 Analyses have shown that this is a cost-effective method of delivering eye care for people with diabetes and is clinically effective in that, in particular settings, it saves more sight through appropriate referral for laser photocoagulation compared with traditional dilated eye examinations performed by ophthalmologists.51 Other studies have shown that implementation of this system increases access to eye care,52

adherence to subsequent eye care,26 subsequent regular clinic visits and improves clinical performance measures.53 A VALIDATED TELEOPHTHALMOLOGICAL PROGRAM

Accurate ocular diagnosis A teleophthalmological program that has undergone rigorous validation and clinical implementation is the Joslin Vision Network (JVN) program, which was established by Bursell in 1998. One of the central JVN studies examined the correlation between the accepted standard of retinal imaging (ETDRS seven standard field 35 mm stereoscopic colour fundus photographs obtained with pupil dilation) and JVN’s ability to diagnose diabetic retinopathy, provide appropriate decision support for follow up and determine the need for referral of the patient to a specialist ophthalmologist.18 For this correlation study, 54 patients (108 eyes) with type 1 or type 2 diabetes were selected through chart review and recruited to receive a JVN examination and an ETDRS imaging protocol. The JVN examination consisted of acquisition of three non-mydriatic 45-degree stereo-retinal fields (posterior pole, superior temporal and nasal) and a single external image to document lens transparency and pupil dilation (affecting image quality), as well as documenting any problems with the iris, cornea, eyelids and conjunctiva (Figure 1). Two independent, masked readers graded the images for both exams, lesion by lesion. An independent ophthalmologist with expertise in the retina adjudicated inter-reader disagreements. The results indicate that the level of agreement between the JVN examination and the ETDRS imaging protocol was high (kappa = 0.65). Additionally, for the assessment of clinically significant macular oedema (the major cause for moderate visual loss) the level of agreement between the assessment from the non-dilated JVN images and the dilated-pupil 35 mm photography was substantial with a kappa statistic of 0.74. Additionally, the level of agreement between the recommended

© 2012 The Authors Clinical and Experimental Optometry © 2012 Optometrists Association Australia

follow up for the JVN examinations and the ETDRS imaging protocol with respect to diabetic retinopathy and macular oedema was high with kappa values of 0.67 and 0.70, respectively. The data also showed that the ability of the JVN system to detect sight-threatening retinopathy was substantial (kappa = 0.88 with sensitivity = 1.0, specificity = 0.86 with a positive predictive value of 0.92 and a negative predictive value of 1.0).

An effective clinical pathway A separate, later study compared a comprehensive dilated eye examination performed by a retinal specialist and the JVN system’s ability to assess the clinical level of diabetic retinopathy and macular oedema, determine follow up and appropriately refer patients to specialist ophthalmologists.43 Five hundred and twenty-five diabetic patients participating in a Joslin Diabetes Centre program on education and intensive treatment training underwent retinal imaging and assessment with the JVN system. Non-mydriatic digital video images were graded for clinical level of diabetic retinopathy and diabetic macular oedema. Patients with unreadable images in any imaged field or for any lesion, any level of diabetic macular oedema, severe or worse non-proliferative diabetic retinopathy, proliferative diabetic retinopathy, last comprehensive eye examinations more than 12 months previously or non-diabetic eye finding requiring prompt attention were referred for a comprehensive eye examination with the retinal specialist during the program or within one week of the program’s conclusion (n = 268). Of the 268 patients who received both types of examination, JVN diagnosis of a clinical level of diabetic retinopathy agreed exactly with the clinical findings in 72.5 per cent of eyes or within one level in 89.3 per cent of eyes. JVN patient referral based on the most severe diagnosis in either eye matched retinal specialist recommended follow up in 92.5 per cent of patients. Additionally, 25.9 per cent of JVN imaged patients had nondiabetic ocular abnormalities requiring referral. Further analysis shows that the JVN ability to differentiate between eyes Clinical and Experimental Optometry 95.3 May 2012

315

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

with no diabetic retinopathy and eyes showing any clinically observable retinal lesions was appropriate (sensitivity = 83 per cent, specificity = 94 per cent, positive predictive value = 91 per cent and negative predictive value = 88 per cent), as was the case when considering discrimination between sight-threatening diabetic retinopathy and milder levels of retinopathy (sensitivity = 93 per cent, specificity = 100 per cent, positive predictive value = 100 per cent and negative predictive value = 75 per cent). The JVN teleretinal program has also been successfully deployed in more than 100 clinical sites within the Indian Health Service (IHS) and at each site the JVN has resulted in increased access to diabetic eye care. In the IHS, the impact of the teleretinal program was measured on the rate of surveillance and treatment of diabetic retinopathy in this large, well-defined patient population over a five-year period.54 A computerised patient information system described, on an annual basis, the patient population, the number of patients with diabetes and the proportion of patients with diabetes who received appropriate medical services, as measured against standards of care before and after implementation of the JVN digital retinal imaging system in a primary care setting. A procedure log ascertained the proportion of patients who received laser treatments for diabetic retinopathy per year. The resultant data analysis showed that the rate of annual retinal examinations increased from 50 per cent (95 per cent confidence interval 44 to 56 per cent) to 75 per cent (95 per cent confidence interval 70 to 80 per cent) (c2 = 58.5, trend p = 0.000001), representing a 50 per cent increase in the retinal examination rate. The laser therapy rate rose from 19.6 per 1000 people with diabetes in 1999 to 29.5 per 1000 people in 2003 for a 51 per cent increase in the laser treatment rate. Thus, implementation of the teleretinal program in a primary care setting led to a significant increase in the rate of surveillance of diabetic retinopathy and a proportional increase in the rate of laser treatment for diabetic retinopathy for a large patient population. Clinical and Experimental Optometry 95.3 May 2012

316

Cost-effectiveness Further studies using the JVN teleretinal program have examined the question of whether the teleretinal program produces gains in effectiveness at lower costs. Using economic modelling techniques, the JVN was found to be a dominant strategy compared with conventional clinic-based eye examinations when used to assess proliferative diabetic retinopathy.51 Dominance means that the JVN resulted in better diagnostic and clinical outcomes and did so at lower costs. Specifically, the economic model used literature-based estimates of operating characteristics, clinical outcomes and administrative and epidemiologic data to predict three outcomes. The outcomes of interest were: 1. the number of true positive cases of proliferative diabetic retinopathy detected 2. the number of patients treated with pan-retinal laser photocoagulation 3. the number of cases of severe visual loss averted. These examination modalities were modelled within the entire diabetic population of the IHS, the Department of Defense, and the Veterans Health Affairs system (VA). The base-case analysis showed that JVN was the dominant strategy in all but one of the 12 modelled scenarios. For example, in the VA, exclusive use of the JVN instead of clinic-based eye examinations would result in 10 fewer cases of severe visual loss per year and the annual cost of screening the large VA population for diabetic retinopathy would decrease by US$4,500,000. The significant cost savings were impressive because at the time of the analysis in the year 2000, the costs for a server to support teleretinal programs was US$400,000 and the cost of a workstation to support the retinal imaging was US$100,000 per workstation. Today an equivalent server costs US$11,000 and an equivalent workstation costs less than US$1,000.

Improved clinical outcomes In a later retrospective study, the relationship of participation in a diabetes telehealth eye-care program to standard,

face-to-face eye care and whether there were also improvements in other diabetes-related health outcomes53 were investigated. The data for the retrospective study included 13,752 patients from the Joslin Diabetes Centre electronic medical record system with a two-year period between baseline and follow up. Subjects were grouped according to eye care, namely, no eye care, eye care outside the clinic and eye care through participation in the JVN teleretinal program, and data were analysed to examine the relationship between participation in the teleretinal program at baseline and follow up eye care, changes in haemoglobin A1c (HbA1c), low-density lipoprotein (LDL) cholesterol levels and systolic blood pressure. The results showed that participation in the telehealth eye-care program was significantly correlated with whether subjects later obtained standard eye care and improvement in HbA1c and LDL cholesterol levels. These data indicate that telehealth eye-care programs that incorporate evaluation, education and care planning are related to improved adherence to recommended eye care and improvements in certain diabetes-related health outcomes. Such programs can address the many aspects of care necessary to reduce the risk of visual loss due to diabetic retinopathy and other diabetes-related complications. In summary, telehealth eye-care programs that combine evaluation, education and care planning that considers the management of systemic diabetes and the effect of risk factors on the development and progression of diabetic retinopathy, can significantly improve health outcomes for patients with diabetes. The act of taking retinal images and treating significant lesions saves vision but can also improve self-management behaviour and related risk factor control, further reducing the risk of visual loss. Thus, appropriately designed and validated telehealth programs that include telehealth eye care can address the many aspects of care necessary to improve clinical outcomes and reduce visual loss and other complications. © 2012 The Authors

Clinical and Experimental Optometry © 2012 Optometrists Association Australia

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

MANAGEMENT OF DIABETES ‘Standard’ chronic disease care typically involves individual health-care professionals reacting to patient-initiated complaints and visits. Care is often fragmented, disorganised, duplicated and focused on managing established disease. Disease management is provider, not patient, directed and focuses on drug and technological interventions with little attention to patient self-management behaviour and physician or health-care professional– patient interactions. Indeed, studies show that people with diabetes are not receiving care known to be beneficial.55,56 In one study, 30 per cent of patients did not have the recommended number of HbA1c or lipid tests.57 In contrast, recent studies from Australia indicate positive results with respect to quality of care and its delivery to Aboriginal communities.58–61 Barriers to care delivery in chronic disease management include low socio-economic status, low health literacy and numeracy and low acculturation.62 Lower socioeconomic status is related to higher prevalence of diabetes and worse diabetesrelated outcomes63 and can impact on selfmanagement. For example, patients with low socio-economic status might find the cost of diabetic supplies, as well as transportation and child-care costs related to clinic attendance and healthier food purchases, difficult. Psychosocial aspects can also impact diabetic care. Patients with diabetes need to manage a wide range of behaviours, including diet and exercise, blood glucose monitoring, hypoglycaemia, drugs, foot care, sick days, check-ups and ongoing education. Complication impacts include treatment demands and co-ordination of GP, specialist and allied health-care clinic visits, onset of symptoms and impaired functioning, and higher financial and time costs. Psychiatric illness and drug and alcohol abuse are common. Family and social pressures, life events and stresses often co-exist. In Australia there is increasing government focus on telehealth64 and Medicare rebates65 for online consultations, financial incentives for service provision, funding

for after-hours consultation, online triage and medical advice via videoconferencing, consideration of optimal practice models and specialties best suited for online consultation. A media release66 detailed telehealth fees and reported that ‘under the Gillard Government’s AUD$620 million telehealth initiative, patients will be able to ‘see’ their specialist close to home without the time and expense of travelling to major cities’. Telehealth will benefit from the National Broadband Network,67 the National Digital Economy Strategy68 and implementation of the Personally Controlled Electronic Health Record.69 These initiatives are important; however, the promise of telehealth improving health outcomes can fall on fallow ground if there is a disconnection between the promise and the implementation. TELEHEALTH AND DIABETES CARE In recent years, there has been a number of publications investigating the use of telehealth applications for the management of diabetes where the data have shown significant improvements in clinical outcomes.60,70–78 None of these applications includes telehealth eye care as part of its function. We have developed a diabetes management and care co-ordination application that includes teleophthalmology as a critical component. This software application, the Chronic Disease Management Program (CDMP), is a web-based open source software application that has undergone research and development over the past 12 years.71,79 The CDMP is a web-based chronic disease management and care co-ordination tool developed by a consortium of researchers, physicians and educators specialising in diabetes and its management. The CDMP development initiative represents many hundreds of programming and research hours that have resulted in the creation of a truly comprehensive disease management program. The open source philosophy encourages software development efforts and sharing the new applications and function across the broader community of users. The clinical function is aligned with the ‘chronic care model’ and develop-

© 2012 The Authors Clinical and Experimental Optometry © 2012 Optometrists Association Australia

ment is based on a ‘service oriented architecture’, where function is grouped around business processes to facilitate interoperable services and data exchange as well as to enhance performance. The CDMP system has been implemented in a number of different environments such as the IHS in over 80 sites across 24 different US states and is interfaced to the IHS electronic health record system (the Resource and Patient Management System). In Australia, we have recently received National Health and Medical Research Council funding to implement the CDMP system including the teleophthalmological module in two Aboriginal-Controlled Community Health Centres in the Northern Territory with an interface to the Communicare electronic health record system. The management of chronic disease, particularly diabetes, requires a combination of enhanced clinical surveillance, lifestyle interventions and medications delivered through a platform of multidisciplinary care provision. In the current health-care system, care is often fragmented, disorganised, redundant and focused on managing established disease and complications. Management of the disease is directed by the provider and focused on pharmacological and technological interventions with little attention given to patient self-management behaviour and provider–patient interactions.80 Educating patients on self-management improves health and decreases use of health-care resources.81–83 The most effective management strategies focus on motivating and more importantly supporting patients to ‘self-manage’ through a team approach involving educators, nutritionists, medical specialists (for example, endocrinologists) and others.84–89 Inadequate access to necessary multidisciplinary care results in preventable death, blindness, amputations and chronic renal failure. Since CDMP was developed as a platform to improve care, it can provide a single application that supports quality improvement across multiple diseases for the primary care provider. CDMP is a clinical data repository that is fully integrated with a number of electronic health Clinical and Experimental Optometry 95.3 May 2012

317

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

records (EHR) and also provides vital disease registry function for monitoring population disease and health service level data for tracking performance over time. It is designed to complement the functions of electronic health records to improve quality and efficiency of healthcare delivery without adding to the burden of record keeping or data analysis. It has the following characteristics: 1. CDMP reorganises clinical data with a snapshot or dashboard page and graphing functions that present data to the provider in a fashion that assists clinical decision-making 2. it has computer-generated decision support with automated reminders to patients and providers for routine procedures and laboratory tests 3. it provides a disease risk assessment engine as an aid to prioritising care 4. it provides a non-transactional anonymous clinical data repository 5. it provides disease registry functions 6. it integrates home monitoring data and other interactions related to the patient’s health back into the medical record 7. a patient portal for personal health record functionality and other personal health record applications including a prototype focused on providing lifestyle decision support for people with diabetes and the ability to accommodate a patient-centred medical home model (the patient portal will be used as the vehicle for content delivery in assisting health workers) 8. the CDMP system has remote e-mail and a secure e-mail system between provider and patient 9. it provides a tool for assessing quality with respect to optimisation of clinical workflow and health-care delivery as well as a tool for identifying quality improvement interventions. Specifically, the CDMP supports health care teams by incorporating current clinical guidelines into decision support algorithms to make the job of managing chronic disease easier for the practitioner and patient. The CDMP was designed to contribute to the standard clinical process by: Clinical and Experimental Optometry 95.3 May 2012

318

1. generating red, yellow and green alerts, based on risk assessment algorithms and recognised management guidelines to direct intensity and elements of evidence-based care 2. providing clinical assessment reminders, notification and communication tools 3. providing an assessment of patient barriers to self-care (including social support, finances, mood et cetera) 4. tracking patient use of educational resources 5. summarising patient knowledge and the impact of educational interventions 6. providing dynamic care planning, which is done with the patient and targets physical wellness, lifestyle selfmanagement and psychosocial health (incorporating information from the clinical and self-care assessments) 7. connecting with the health organisation’s health information system or available electronic data (with provisions for client and medical records privacy). The CDMP provides patients with direct access to their personal health information, standardised learning materials offered by their health-care team (with tools to track patient compliance and evaluation of the impact of the educational intervention on patient behaviour) and care team communications. Key components include: 1. biometric (for example, blood glucose test) data upload with automated feedback 2. personal care plans created in concert with the patients care manager 3. a personalised learning plan using preapproved training materials 4. access to data including laboratory results, medications, procedures and diagnoses 5. secure patient/provider electronic private messaging/communications 6. automated patient notification when new data or analyses are available 7. access to public discussions on diabetes and support forums. The CDMP integration into the Australian primary care context will include dynamic support of health workers and

other clinic staff to make the most of the programs function. It is expected that CDMP usage will: 1. significantly improve health practitioner confidence and capacity in managing chronic disease in indigenous communities 2. improve the delivery of evidence-based care for diabetes and cardiovascular disease 3. improve health practitioner skills and resources within primary care centres to support patients to better selfmanage their chronic conditions 4. improve patient confidence and skills in self-managing chronic conditions 5. provide a template for deployment of system level improvements in chronic disease care through the CDMP in geographically dispersed and underserved populations. The CDMP has two user interfaces: the first for providers and care managers (referred to as CDMP) and the second for direct patient access (referred to as DMEverywhere). The CDMP provides an automatic system to foster continuous care and secure communication among patients and providers, ensures that the latest clinical guidelines are applied in the care and focuses on clinical outcomes, education and patients’ barriers and successes in their self-care. The CDMP care model provides critical health-care co-ordination between patients and the health-care team including medical practitioners, nurse practitioners, diabetes educators, sub-specialists (such as, optometrists and podiatrists), dietitians/nutritionists, behavioural clinicians, pharmacists and extension programs such as Aboriginal Health Workers (AHW). The common theme within the CDMP application is the ability to facilitate connectivity and linkage among various members of an extended care team including the AHW and the patient, and this theme is the key to understanding where CDMP fits in the picture of health information technology for its use in online care programs. CDMP is designed around a team model. It has role-defined sign-on that can present role-defined views of the clinical © 2012 The Authors

Clinical and Experimental Optometry © 2012 Optometrists Association Australia

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

information and a role-limited ability to control information and alerts. More significant is the ability to have a structured system of documentation for roles that have specific services. Thus, the dietitian has pull-down menus and assessment tools specifically for his/her role. The role definition will need to be expanded for health centre services and the broadly defined care team but the CDMP platform is designed to incorporate these definitions. The ‘comprehensive health home’ focuses on integrating care at the health centre with the larger community and thereby facilitates care co-ordination. CDMP easily pushes information out to community partners and the information can be accessed over the Internet. The role-based access in CDMP allows for limiting the information viewed by community partners to a ‘need-to-know’ basis, thereby addressing some of the privacy issues that arise in sharing data beyond organisational walls. Through its patient portal function, CDMP provides patients with enhanced access to their health-care team, as well as the opportunity for greater participation in their own care. Through the portal’s various patient feedback mechanisms and training materials that are incorporated into the CDMP library, the patient is provided with self-management support tools that augment the support provided by the care team of the ‘comprehensive health home’. Currently, CDMP has robust decision support for management of diabetes and uses a sophisticated decision support system that accounts for the patient’s current condition in generating reminders and alerts. Decision support for cardiovascular and liver disease has also been built into the system. While it is recognised that decision support requires a high level of clinical input to create and maintain, the system infrastructure is designed to easily incorporate complex decision support algorithms and could be further developed to include additional conditions. Figure 2 shows the CDMP ‘dashboard’ patient summary. The tabs on the left side show the applications of the CDMP. The content of these tabs can be modified to meet the

Figure 2. Patient dashboard screen from the telehealth Chronic Disease Management Program (CDMP)

requirements of any program. Of interest is the ‘white board’ panel on the right that facilitates communication on patient issues that can be entered at the time of the encounter. This is especially useful for Aboriginal health workers as a way of

© 2012 The Authors Clinical and Experimental Optometry © 2012 Optometrists Association Australia

rapidly communicating patient issues with the rest of the care team. Finally, CDMP has a built-in open source report writer along with a library of pre-defined reports as illustrated in Figure 3. This function includes the Clinical and Experimental Optometry 95.3 May 2012

319

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

Thus, telehealth programs that incorporate evaluation, education and care planning are associated with improved clinical outcomes and self-management behaviour. Such programs can address the many aspects of care and care co-ordination necessary to reduce the risk of diabetic complications, including visual loss due to diabetic retinopathy. A PROPOSED TELEOPHTHALMOLOGICAL WORKFLOW MODEL FOR PEOPLE WITH DIABETES

Figure 3. An example of a CDMP encounter report form screen

population-based reporting required to support the comprehensive medical home. The flexibility in the reporting engine also provides for performance reporting on a sub-population or provider level that facilitates a rapid identification of gaps in care, potential quality improvement interventions and could reduce resources required by clinics and health centres to manually generate reports. In a recently published study,79 results using the CDMP application, including Clinical and Experimental Optometry 95.3 May 2012

320

the use of the telehealth eye-care module, demonstrated significant improvement in clinical outcomes for diabetics. Furthermore, the results showed that the improvement after six months was sustained for at least one year, the longest follow up. In addition to the improved clinical outcomes, there was a significant reduction in the level of stress patient’s experienced in managing their own diabetes measured using the Problem Areas In Diabetes tool.

Eligible patients for teleophthalmology are people with diabetes, who have not accessed standard eye care with either an optometrist or ophthalmologist in the previous two years (one year if there is a high risk for complications, such as being of indigenous ethnicity), are generally situated in remote or under-resourced areas and are not visually impaired (individuals with visual impairment require comprehensive eye examinations). In this model, teleophthalmology is not a broad-based national screening program. The service is incorporated into the usual cycle of chronic disease management for eligible patients. Figure 4 illustrates the professional groups involved in this teleophthalmology-based diabetes cycle of care. That is, teleophthalmology is requested, performed or discussed during a routine medical consultation, field trip or service (for example, blood test) related to the regular cycle of care of people with diabetes. Medical practitioners or optometrists identify eligible patients with diagnosed diabetes (that is, those not visually impaired and not examined within the previous two years by a registered eye-care practitioner). They either perform nonmydriatic retinal photography (NMP) or authorise accredited imagers to do so under their supervision. Presenting distance vision in each eye is recorded as part of the imaging procedure to detect impaired vision of recent onset. A broadly defined imaging protocol for the detection of diabetic retinopathy is as follows: © 2012 The Authors

Clinical and Experimental Optometry © 2012 Optometrists Association Australia

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

Patient attends health check

Diabetes diagnosis

No

Yes

Medical practitioner GP Diabetologist or other

Diabetes management: Blood/urine

Pathologist Dietitian Podiatrist Optometrist or ophthalmologist

Diet Feet Eyes

Retinal photography

Trained technician Trained nurse Trained health worker Optometrist Medical practitioner

Reading retinal photos

Eye health report

Trained medical practitioner Optometrist Accredited reader

Figure 4. Professional groups involved in the teleophthalmology-based diabetes cycle of care

1. photograph(s) of each eye should be centred on the macula and include the optic disc and major vascular arcades 2. photograph(s) can be taken with either a non-mydriatic or a mydriatic retinal camera 3. stereoscopic images are not required. Technical failure can occur with any digital device, including NMP, that is, failure to obtain an image of adequate quality for grading. Technical failure is relatively low in NMP since excluding visual impairment eliminates those with substantial cataracts or vitreous haemorrhage, thereby eliminating common causes for technical failure. The grading protocol also grades image quality. One protocol for grading of image quality is as follows: 1. nerve fibre layer is visible 2. nerve fibre layer is not visible 3. small vessels are blurred 4. major arcade vessels just blurred

5. significant blurring of major arcade vessels in more than one-third of the image. Degraded images are generally due to cataract or small pupils. The grade of image quality that should be considered to be a technical failure needs to be standardised. Accredited readers report on the quality of the images, the level of presenting vision in each eye, the level of diabetic retinopathy and indicated referral timeframe (where necessary) to the health professional, who authorised the service and who, in turn, refers the patient for a comprehensive eye examination, if diabetic retinopathy or visual impairment is detected. A referred patient moves to standard care and becomes ineligible for subsequent NMP, unless eligibility criteria are met in the future. Reading of retinal images can be performed by an ophthalmologist, optometrist or any other Medicare service

© 2012 The Authors Clinical and Experimental Optometry © 2012 Optometrists Association Australia

provider who has appropriate training. For example, recognition of the features of diabetic retinopathy has been an integral component of the upskilling of general practitioners in recent years. It is easier for general practitioners to recognise early retinopathy on a digital image than through an ophthalmoscope as they are currently expected to do, especially through an undilated pupil. Formal grader training of general practitioners interested in retinal grading could be provided by ophthalmologists or optometrists. Other forms of training using a dedicated web-based accreditation could also be considered. Regardless of the training opportunities, the effectiveness of the provision of these services depends on the resources used to provide them. Thus, in some circumstances it might not be the best use of general practitioner time to be trained for and to provide these ocular services. As part of the grading process, abnormalities detected in retinal images or reduction in vision in either eye should prompt a referral for evaluation and management to an eye-care professional. The report is generated and sent by the reader to both the imaging centre and the referring practitioner within a specified time. Upon discussing the report with the patient, the practitioner either refers the patient for a comprehensive eye examination (for evaluation of detected retinopathy/visual impairment or if images were ungradable) or refers for repeat NMP in two years (one year, if indigenous), provided the patient continues to meet the inclusion criteria for NMP. 1. For sustainability of NMP, a Medicare item number would be available as part of the overall health management of the Australian population with diabetes. 2. NMP would be authorised by a medical practitioner or optometrist and provided by a medical practitioner, optometrist or accredited staff such as a technician, nurse or health worker and an accredited reader. The reader could be a medical practitioner, optometrist or accredited staff at an accredited reading centre. If optometrists or Clinical and Experimental Optometry 95.3 May 2012

321

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

DR severity level

Findings observable on dilated retinal examination

No apparent DR

No abnormalities

Mild NPDR

Microaneurysms only

Moderate NPDR

More than just microaneurysms but less than severe NPDR

Severe NPDR

Any of the following (4-2-1 rule) and no signs of proliferative retinopathy: Severe intraretinal haemorrhages and microaneurysms in each of four quadrants Definite venous beading in two or more quadrants Moderate IRMA in one or more quadrants

PDR

One or both of the following: Neovascularisation Vitreous/pre-retinal haemorrhage

DR: diabetic retinopathy, IRMA: intraretinal microvascular abnormalities, NPDR: non-proliferative diabetic retinopathy, PDR: proliferative diabetic retinopathy

Table 2. Severity scale for diabetic retinopathy

ophthalmologists perform the reading, then these parties would obtain a portion of the fee from the Medicare service provider who authorised the NMP service. 3. The red colour in Figure 4 indicates the services and providers involved in this teleophthalmological model. CLINICAL GRADING OF DIABETIC RETINOPATHY The development and implementation of national guidelines for evidence-based clinical practice can help ensure that clinical practice is up to date and consistent with internationally recognised best practice. Consequently, clinical grading of diabetic retinopathy in Australia is based on the severity scale for diabetic retinopathy (Table 2), as found in the 2008 Clinical Practice Guidelines for the Diagnosis, Management and Prevention of Diabetic Retinopathy.13 The objective of these guidelines is to assist practitioners in making decisions about the appropriate health care of patients with diabetes and to inform health-care professionals and consumers about the need for regular eye examinations to detect diabetic retinopathy in a timely manner to enable successful treatment and preserve vision. Clinical and Experimental Optometry 95.3 May 2012

322

VISION AND CHALLENGES FOR THE FUTURE Based on overseas experience in the US and in the UK, the implementation and long-term sustainment of a clinically effective teleophthalmological program is possible and desirable, particularly in countries with a geographically dispersed population and unevenly distributed skilled health-care workforce, such as Australia. For teleretinal/teleophthalmological programs to be truly effective in the telehealth environment there remains a number of technological challenges to be met, including better telehealth retinal cameras, advanced software applications for automated diagnosis of the level of clinical disease with the capability to support predictive decisions, internet and wireless connectivity, staffing resource issues around changing clinical workflows, staff training in telehealth applications and appropriate funding and/or reimbursement for these services.

Retinal camera technology The telehealth retinal camera system needs to be small, light weight and rugged enough to travel to regions with extreme (hot, cold or dusty) environments, flexible enough to be transported under various

modes of travel, such as light planes or dog sleds and be simple enough to be operated and supported by local healthcare workers. Currently, the available portable cameras include the Kowa and Nidek systems, but for telehealth requirements these commercial systems still need further optimisation for use under remote/extreme environmental conditions and to be available at lower costs, so they are affordable for smaller practices. Current, commercially available nonmydriatic retinal imaging systems including Topcon, Canon, Kowa and Nidek tend to be expensive, not very portable and their use is limited to pupil diameters no less than 3.5 mm. Smaller pupils significantly impact on image quality. Additionally, the levels of light used to expose the retinal images tend to be high and uncomfortable for the patient. A true telehealth device for retinal imaging needs to be inexpensive, portable, wireless capable for image transmission, capable of imaging the retina through smaller pupils and must use lower light levels for exposure of retinal images. Thus, the system becomes less intrusive and fosters increased access to and improved adherence with appropriate eye care. There are less expensive and portable retinal cameras in development; however, the ideal would be an adaptor to the smartphone that allows retinal imaging of sufficiently high quality to enable accurate diagnoses of clinical retinal disease.

Advanced software applications for automated image assessment In addition to the need for technological advances in this arena, there is also a need and potential to improve the assessment of retinal disease through the support of computer algorithms. Currently, the required retinal assessment is still performed manually by eye-care professionals. This is a high cost that can be significantly reduced through the use of real-time computer-assisted diagnostic algorithms and treatment plans for appropriate eye care. For these algorithms to have any degree of clinical acceptance, the sensitivity and specificity for diagnosis of the level of diabetic retinopathy needs to © 2012 The Authors

Clinical and Experimental Optometry © 2012 Optometrists Association Australia

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

be in the 90 per cent range. Thus, there is a critical need for improved computer algorithms for automated assessment of the level of retinopathy. In reviewing the literature on retinal image analysis, most progress is in the automated grading process with the goal of replacing the high resources used in current manual grading. In general, the use of computers to aid the detection of retinopathy has been accomplished in as many different ways as there are groups performing the work.90–101 The use of neural networks that try and ‘teach’ computing systems to recognise patterns has not yet achieved a high enough sensitivity and specificity to be used The simplest algorithms are those that differentiate between a disease and a no disease state. The determination of specific types of retinopathy has been attempted by a number of groups with varying levels of sensitivity and specificity for detecting specific retinal lesions. Techniques to extract features are also being developed that allow the tracing of retinal vessels and crossings (where retinal vessels cross each other) using regiongrowing and edge-detection algorithms. Feature extraction is a fundamental step in a fully automated analysing system, because the computer needs a mathematically exact definition of the data to be analysed. These techniques allow the registration and fusing of retinal images but retinal image registration is complicated. The images are of a curved surface taken from a wide range of viewpoints sometimes using an uncalibrated camera with artefacts arising from reflection, refraction and dispersion. Involuntary body and eye movements compound these effects. A novel approach, namely content-based image retrieval, has been applied recently to computer-based analysis of retinal images.102,103 The concept is based on a process of retrieving related images from a large database of adjudicated retinal images and using their pictorial content to provide the correct assessment of clinical diagnosis. The resultant list of features provides the index for storage, search and retrieval of related images from the library based on their specific visual characteris-

tics. The probabilistic nature of contentbased image retrieval allows for statistically appropriate predictions on presence, severity and manifestations of common retinal diseases, such as diabetic retinopathy. This is an attractive approach but might suffer performance issues, because the library could become unwieldy given the diversity of characteristics and spatial locations of these retinal lesions. The common thread and the resultant limited success are probably related to a focus on the exact definition of a retinal lesion. This is an extremely difficult process for a computer to learn; however, it is not what an eye-care professional goes through when performing a retinal evaluation and providing a diagnosis. The eyecare professional will generally look for a constellation of lesions, real or not, using a cognitive learning process to arrive at an accurate diagnosis. This is a pattern recognition process that not only tries to identify lesions but also the relationship between the various lesions. Thus, the exact identification of any one lesion is of lesser importance for diagnosis than the potentially different types of lesions and where they are located. Thus, a ‘missing’ lesion becomes less important to the end diagnosis. Methodologies developed by armed forces for detection of military targets that use pattern recognition technologies with cognitive learning for computer-assisted diagnosis can be leveraged for detecting patterns in diabetic retinopathy. To our knowledge, its application to retinal image analysis is currently untried; however, the methodology is attractive, because it closely approximates what an eye-care professional does when assessing the retina and technologically can achieve high performance rates.

Internet connectivity for telehealth and retinal image transmission The challenges to be resolved include the availability, affordability and reliability of broadband internet connectivity. Telecommunications, even in a nationally funded broadband roll-out are not always provided to the more remote areas, where gaps in care and need for health-care ser-

© 2012 The Authors Clinical and Experimental Optometry © 2012 Optometrists Association Australia

vices are greatest, and if they are available, they might not be consistently reliable. Alternatives for retinal image transmission are the use of landline telephone networks, regular postage mailing of CDs or USB memory devices or by providing local capabilities, such as automated diagnostics through computer algorithms that can be adjudicated at later times by a geographically remote expert. These alternatives are a poor substitute for dynamic telehealthcare service delivery.

Human resources Staff turnover can be an issue, particularly in remote areas. Simple to operate retinal cameras with clear written instructions and online training and education modules should be provided, so that training and cross-training of retinal imagers is feasible. Turnover of retinal grading centre staff is usually less of an issue, particularly if several graders or several reading centres are available. In the telehealth arena there needs to be a focus on capacity building that includes the training of telehealthcare nurses and managers and the training of Australian Health Workers in the optimal use of telehealth-care delivery.

Incorporation of clinically relevant data To achieve high value and clinical effectiveness for diabetes management, teleretinal and teleophthalmological programs must become a part of the telehealth services available to patients. This represents a more holistic approach. Teleretinal management plans must incorporate the patient’s other medical data, which as well as clinical measures and laboratory data should include data from remote monitoring devices in the home and smartphone remote applications for monitoring from anywhere. The current environment is moving to the ubiquitous use of smartphones for health. To appropriately leverage this movement, there needs to be multidisciplinary collaborative efforts that include physicians, clinical scientists, health-care research scientists, social scientists, behavioural scientists, public health scientists, psychologists and health informatics scientists with a focus on investigatClinical and Experimental Optometry 95.3 May 2012

323

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

ing the optimum health impact of integrated care through the use of mobile health technologies and the role of social support networks for patients and health providers in health-care delivery.

Adherence People with health problems often do not adhere to medically prescribed health regimens that could reduce or prevent longterm chronic diseases, even if they and the health-care system can afford the skyrocketing medical costs. Adherence to recommended diabetic care, for example, could save US$62 million to US$109 million per year in US health-care costs. Self-management of life-threatening diseases (for example, diabetes, heart disease, human immunodeficiency virus and related mental health problems) is a major public health issue globally. Patients with chronic diseases, such as diabetes, live with the disease every minute of every day and new technological innovations for healthcare management can significantly improve patient health and wellness by improving patient self-management behaviour and reducing their level of stress in dealing with their disease. The reduction in chronic disease risk will result in a reduction in the occurrence of complications, such as diabetic retinopathy and blindness. Traditional methods of health-care delivery are increasingly being augmented by adoption of merging electronic health records and health information exchange technologies that focus on remote patient monitoring, telehealth and mobile health. These technologies will be vital to meet global demands for efficient, clinically effective and affordable health-care delivery. There is an urgent public health need to harness and adapt the newest mobile health technologies for expanding the reach and delivery of health-care services. This will allow patients, through a simple and central place, to connect with each other and their health-care team (including friends, family, physicians, care teams, education coaches, peer mentors, health plans, pharmacies, hospitals and employers) to improve their health through these interactions. The adoption of mobile Clinical and Experimental Optometry 95.3 May 2012

324

health as a solution for health-care service delivery is expanding rapidly and will enable and empower underserved populations to manage their health problems. Thus, adoption of mobile health technologies will result in improved health-care access, socially empowered behavioural changes in self-management to promote wellness and reduce complications such as visual loss and reduce health disparities globally.

Payment Who pays for health care is always a challenge. Existent models include a mix of government funded, private health insurance and self-pay. The pharmaceutical and device (for example, insulin pump) industry also contribute. Emerging mobile health technologies tend to rely on user subscription payments. Current teleophthalmological programs are funded by the Government and by private health insurance companies. In Australia, many people already pay for retinal photography performed by their optometrist and hence the service is already seen by many as worthwhile. Cost-sharing between Government, private health insurance, industry and the consumer is likely to be sustainable. As the program is adopted and the related technology improves, including automated or semi-automated diagnosis, costs will become reduced. SUMMARY AND CONCLUSION In the past two decades, the expansion of information technology infrastructure, mobile computing, enhanced telecommunications and electronic epidemiological and clinical information systems has revolutionised the way in which health care is delivered to patients across the continuum of care. Within this space, telehealth/ telemedicine offers important opportunities for improving the delivery and outcomes of evidence-based care to vulnerable and geographically isolated individuals and those experiencing barriers to accessing necessary care. Despite the promise that telehealth and telemedicine holds, adoption in Australia falls behind

other developed and developing countries, and as a consequence the benefits of mobile technological advances in healthcare delivery are yet to be fully realised. Furthermore, the use of this telehealth infrastructure might also offer significant benefit to the overall health of disadvantaged and resource-poor communities across the globe. Teleretinal/teleophthalmological surveillance costs less and in some settings saves more sight than ophthalmologistbased dilated eye examinations (even in the costly US health-care system) and leads to improved adherence to subsequent eye care, regular clinical appointments and improvements in major vascular risk factors, such as hyperglycaemia and elevated lipid levels. Cost-effective and sustainable teleretinal surveillance to facilitate the management of diabetic retinopathy and other eye diseases requires the combination of an affordable portable device for taking low light-level retinal images (a data collection device) without the use of pharmacological pupil dilation, a computer-assisted methodology for rapidly detecting and diagnosing diabetic retinopathy (a data analysis application) and other sight-threatening ocular disorders and an integrating information technology platform for co-ordinating multidisciplinary evidence-based care. The development of these critical platforms has significant implications for reducing preventable blindness and related vascular disease in developing countries and in marginalised populations (such as indigenous Australians) within high-income countries. By using ocular images as a critical lever, supported by integrated health information technology, telehealth-care service delivery systems are able to focus the attention of patients and their multitude of providers on the eye health of patients and, as a consequence, have seen significant improvements in chronic disease management and outcomes. ACKNOWLEDGEMENTS

The authors acknowledge the helpful discussions with colleagues, including Dr Kevin Rowley, Dr Alex Brown, Professor © 2012 The Authors

Clinical and Experimental Optometry © 2012 Optometrists Association Australia

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

Hugh Taylor, Associate Professor David O’Neal, Associate Professor Ecosse Lamoureux and Professor Tien Wong.

12.

GRANTS AND FINANCIAL ASSISTANCE

The authors’ involvement in this clinical and research area in Australia is supported by a National Health and Medical Research Council Partnership Project grant APP1016691 and by the Fred Hollows Foundation.

13.

14.

REFERENCES 1. Whiting DR, Guariguata L, Weil C, Shaw J. IDF diabetes atlas: global estimates of the prevalence of diabetes for 2011 and 2030. Diabetes Res Clin Pract 2011; 94: 311–321. 2. Australian Institute of Health and Welfare. Diabetes: Australian facts 2008. Cat. no. CVD 40. Canberra: Australian Institute of Health and Welfare, 2008. 3. Wang YC, McPherson K, Marsh T, Gortmaker SL, Brown M. Health and economic burden of the projected obesity trends in the USA and the UK. Lancet 2011; 378: 815–825. 4. Trasande L, Elbel B. The economic burden placed on healthcare systems by childhood obesity. Expert Rev Pharmacoecon Outcomes Res 2012; 12: 39–45. 5. The Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulindependent diabetes mellitus. N Engl J Med 1993; 329: 977–986. 6. Keech AC, Mitchell P, Summanen PA, O’Day J, Davis TM, Moffitt MS, Taskinen MR et al. Effect of fenofibrate on the need for laser treatment for diabetic retinopathy (FIELD study): a randomised controlled trial. Lancet 2007; 370: 1687– 1697. 7. Genuth S, Ismail-Beigi F. Clinical implications of the ACCORD trial. J Clin Endocrinol Metab 2012; 97: 41–48. 8. Fong DS, Strauber SF, Aiello LP, Beck RW, Callanan DG, Danis RP, Davis MD et al. Comparison of the modified Early Treatment Diabetic Retinopathy Study and mild macular grid laser photocoagulation strategies for diabetic macular edema. Arch Ophthalmol 2007; 125: 469– 480. 9. Googe J, Brucker AJ, Bressler NM, Qin H, Aiello LP, Antoszyk A, Beck RW et al. Diabetic Retinopathy Clinical Research Network: randomized trial evaluating short-term effects of intravitreal ranibizumab or triamcinolone acetonide on macular edema after focal/grid laser for diabetic macular edema in eyes also receiving panretinal photocoagulation. Retina 2011; 31: 1009– 1027. 10. Yau JW, Rogers SL, Kawasaki R, Lamoureux EL, Kowalski JW, Bek T, Chen SJ et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care 2012; 35: 556–564. 11. Chow SP, Aiello LM, Cavallerano JD, Katalinic P, Hock K, Tolson A, Kirby R et al. Comparison of nonmydriatic digital retinal imaging versus

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

dilated ophthalmic examination for nondiabetic eye disease in persons with diabetes. Ophthalmology 2006; 113: 833–840. Fong DS, Aiello L, Gardner TW, King GL, Blankenship G, Cavallerano JD, Ferris FL 3rd et al. Retinopathy in diabetes. Diabetes Care 2004; 27 (Suppl 1): S84–S87. Australian Government National Health and Medical Research Council NHMRC. Guidelines for the management of diabetic retinopathy. 2008. Available from: http://www.nhmrc.gov. au/_files_nhmrc/publications/attachments/ di15.pdf?q=publications/synopses/_files/ di15.pdf. Accessed February 2012. McCarty CA, Lloyd-Smith CW, Lee SE, Livingston PM, Stanislavsky YL, Taylor HR. Use of eye care services by people with diabetes: the Melbourne Visual Impairment Project. Br J Ophthalmol 1998; 82: 410–414. McCarty DJ, Fu CL, Harper CA, Taylor HR, McCarty CA. Five-year incidence of diabetic retinopathy in the Melbourne Visual Impairment Project. Clin Experiment Ophthalmol 2003; 31: 397– 402. Schoenfeld ER, Greene JM, Wu SY, Leske MC. Patterns of adherence to diabetes vision care guidelines: baseline findings from the Diabetic Retinopathy Awareness Program. Ophthalmology 2001; 108: 563–571. Chin MH, Cook S, Jin L, Drum ML, Harrison JF, Koppert J, Thiel F et al. Barriers to providing diabetes care in community health centers. Diabetes Care 2001; 24: 268–274. Bursell SE, Cavallerano JD, Cavallerano AA, Clermont AC, Birkmire-Peters D, Aiello LP, Aiello LM. Stereo nonmydriatic digital-video color retinal imaging compared with Early Treatment Diabetic Retinopathy Study seven standard field 35-mm stereo color photos for determining level of diabetic retinopathy. Ophthalmology 2001; 108: 572–585. Aiello LM. Preserving human vision: eliminating blindness from diabetes mellitus. J Am Optom Assoc 1998; 69: 690–692. Zimmer-Galler I, Zeimer R. Results of implementation of the DigiScope for diabetic retinopathy assessment in the primary care environment. Telemed J E Health 2006; 12: 89–98. Olayiwola JN, Sobieraj DM, Kulowski K, St Hilaire D, Huang JJ. Improving diabetic retinopathy screening through a statewide telemedicine program at a large federally qualified health center. J Health Care Poor Underserved 2011; 22: 804–816. Ahmed J, Ward TP, Bursell SE, Aiello LM, Cavallerano JD, Vigersky RA. The sensitivity and specificity of nonmydriatic digital stereoscopic retinal imaging in detecting diabetic retinopathy. Diabetes Care 2006; 29: 2205–2209. Aspelund T, Thornorisdottir O, Olafsdottir E, Gudmundsdottir A, Einarsdottir AB, Mehlsen J, Einarsson S et al. Individual risk assessment and information technology to optimise screening frequency for diabetic retinopathy. Diabetologia 2011; 54: 2525–2532. Boucher MC, Gresset JA, Angioi K, Olivier S. Effectiveness and safety of screening for diabetic retinopathy with two nonmydriatic digital images compared with the seven standard stereoscopic photographic fields. Can J Ophthalmol 2003; 38: 557–568.

© 2012 The Authors Clinical and Experimental Optometry © 2012 Optometrists Association Australia

25. Bragge P, Gruen RL, Chau M, Forbes A, Taylor HR. Screening for presence or absence of diabetic retinopathy: a meta-analysis. Arch Ophthalmol 2011; 129: 435–444. 26. Conlin PR, Fisch BM, Cavallerano AA, Cavallerano JD, Bursell SE, Aiello LM. Non-mydriatic teleretinal imaging improves adherence to annual eye examinations in patients with diabetes. J Rehabil Res Dev 2006; 43: 733–740. 27. Conlin PR, Fisch BM, Orcutt JC, Hetrick BJ, Darkins AW. Framework for a national teleretinal imaging program to screen for diabetic retinopathy in Veterans Health Administration patients. J Rehabil Res Dev 2006; 43: 741–748. 28. Goatman KA, Philip S, Fleming AD, Harvey RD, Swa KK, Styles C, Black M et al. External quality assurance for image grading in the Scottish Diabetic Retinopathy Screening Programme. Diabet Med. 2011; doi: 10.1111/j.1464-5491.2011. 03504.x 29. Hejlesen O, Ege B, Englmeier KH, Aldington S, McCanna L, Bek T. TOSCA-Imaging–developing Internet based image processing software for screening and diagnosis of diabetic retinopathy. Stud Health Technol Inform 2004; 107: 222–226. 30. Kumar SK, Kanagasingam Y, Constable IJ. Should telemedicine in eye care be funded in Australia? Med J Aust 2004; 181: 583. 31. Luzio S, Hatcher S, Zahlmann G, Mazik L, Morgan M, Liesenfeld B, Bek T et al. Feasibility of using the TOSCA telescreening procedures for diabetic retinopathy. Diabet Med 2004; 21: 1121–1128. 32. Peng J, Zou H, Wang W, Fu J, Shen B, Bai X, Xu X et al. Implementation and first-year screening results of an ocular telehealth system for diabetic retinopathy in China. BMC Health Serv Res 2011; 11: 250. 33. Phiri R, Keeffe JE, Harper CA, Taylor HR. Comparative study of the polaroid and digital nonmydriatic cameras in the detection of referrable diabetic retinopathy in Australia. Diabet Med 2006; 23: 867–872. 34. Ting DS, Tay-Kearney ML, Constable I, Lim L, Preen DB, Kanagasingam Y. Retinal video recording a new way to image and diagnose diabetic retinopathy. Ophthalmology 2011; 118: 1588–1593. 35. Ting DS, Tay-Kearney ML, Kanagasingam Y. Light and portable novel device for diabetic retinopathy screening. Clin Experiment Ophthalmol 2012; 40: e40–e46. 36. National Health Service (NHS) Diabetic Eye Screening Program. UK Screening Portal, 2012. Available from: http://diabeticeye.screening. nhs.uk/. Accessed 17 February 2012. 37. Turner AW, Mulholland W, Taylor HR. Funding models for outreach ophthalmology services. Clin Experiment Ophthalmol 2011; 39: 350–357. 38. Grading diabetic retinopathy from stereoscopic color fundus photographs–an extension of the modified Airlie House classification. ETDRS report number 10. Early Treatment Diabetic Retinopathy Study Research Group. Ophthalmology 1991; 98 (5 Suppl): 786–806. 39. Wilkinson CP, Ferris FL 3rd, Klein RE, Lee PP, Agardh CD, Davis M, Dills D et al. Proposed international clinical diabetic retinopathy and diabetic macular edema disease severity scales. Ophthalmology 2003; 110: 1677–1682. 40. Fransen SR, Leonard-Martin TC, Feuer WJ, Hildebrand PL. Clinical evaluation of patients

Clinical and Experimental Optometry 95.3 May 2012

325

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

41.

42.

43.

44.

45.

46.

47.

48.

49. 50.

51.

52.

53.

54.

55.

with diabetic retinopathy: accuracy of the Inoveon diabetic retinopathy-3DT system. Ophthalmology 2002; 109: 595–601. Rudnisky CJ, Hinz BJ, Tennant MT, de Leon AR, Greve MD. High-resolution stereoscopic digital fundus photography versus contact lens biomicroscopy for the detection of clinically significant macular edema. Ophthalmology 2002; 109: 267– 274. Schneider S, Aldington SJ, Kohner EM, Luzio S, Owens DR, Schmidt V, Schuell H et al. Quality assurance for diabetic retinopathy telescreening. Diabet Med 2005; 22: 794–802. Cavallerano AA, Cavallerano JD, Katalinic P, Tolson AM, Aiello LP, Aiello LM. Use of Joslin Vision Network digital-video nonmydriatic retinal imaging to assess diabetic retinopathy in a clinical program. Retina 2003; 23: 215–223. Murgatroyd H, Ellingford A, Cox A, Binnie M, Ellis JD, MacEwen CJ, Leese GP. Effect of mydriasis and different field strategies on digital image screening of diabetic eye disease. Br J Ophthalmol 2004; 88: 920–924. Scanlon PH, Foy C, Malhotra R, Aldington SJ. The influence of age, duration of diabetes, cataract, and pupil size on image quality in digital photographic retinal screening. Diabetes Care 2005; 28: 2448–2453. Li HK, Horton M, Bursell SE, Cavallerano J, Zimmer-Galler I, Tennant M, Abramoff M et al. Telehealth practice recommendations for diabetic retinopathy, second edition. Telemed J E Health 2011; 17: 814–837. Stanberry B. The legal and ethical aspects of telemedicine. 3: Telemedicine and malpractice. J Telemed Telecare 1998; 4: 72–79. Iserson KV. Telemedicine: a proposal for an ethical code. Camb Q Healthc Ethics 2000; 9: 404– 406. Mossman KL. Medical testing: issues and ethics. Forum Appl Res Public Policy 1997; 12: 90–101. Silverman RD. Current legal and ethical concerns in telemedicine and e-medicine. J Telemed Telecare 2003; 9 (Suppl 1): S67–S69. Whited JD, Datta SK, Aiello LM, Aiello LP, Cavallerano JD, Conlin PR, Horton MB et al. A modeled economic analysis of a digital teleophthalmology system as used by three federal health care agencies for detecting proliferative diabetic retinopathy. Telemed J E Health 2005; 11: 641–651. Ng M, Nathoo N, Rudnisky CJ, Tennant MT. Improving access to eye care: teleophthalmology in Alberta, Canada. J Diabetes Sci Technol 2009; 3: 289–296. Fonda SJ, Bursell SE, Lewis DG, Garren J, Hock K, Cavallerano J. The relationship of a diabetes telehealth eye care program to standard eye care and change in diabetes health outcomes. Telemed J E Health 2007; 13: 635–644. Wilson C, Horton M, Cavallerano J, Aiello LM. Addition of primary care-based retinal imaging technology to an existing eye care professional referral program increased the rate of surveillance and treatment of diabetic retinopathy. Diabetes Care 2005; 28: 318–322. Arday DR, Fleming BB, Keller DK, Pendergrass PW, Vaughn RJ, Turpin JM, Nicewander DA. Variation in diabetes care among states: do patient characteristics matter? Diabetes care 2002; 25: 2230–2237.

Clinical and Experimental Optometry 95.3 May 2012

326

56. Chin MH, Auerbach SB, Cook S, Harrison JF, Koppert J, Jin L, Thiel F et al. Quality of diabetes care in community health centers. Am J Public Health 2000; 90: 431–434. 57. Phillips P. Diabetes in Australia—our 2009 report card. Diabetes Australia. Conquest 2010; 10–11. 58. Bailie R, Si D, Connors C, Weeramanthri T, Clark L, Dowden M, O’Donohue L et al. Study protocol: Audit and Best Practice for Chronic Disease Extension (ABCDE) Project. BMC Health Serv Res 2008; 8: 184. 59. McDermott RA, McCulloch BG, Campbell SK, Young DM. Diabetes in the Torres Strait Islands of Australia: better clinical systems but significant increase in weight and other risk conditions among adults, 1999–2005. Med J Aust 2007; 186: 505–508. 60. Rowley KG, O’Dea K, Anderson I, McDermott R, Saraswati K, Tilmouth R, Roberts I et al. Lower than expected morbidity and mortality for an Australian Aboriginal population: 10-year follow-up in a decentralised community. Med J Aust 2008; 188: 283–287. 61. Valery PC, Moloney A, Cotterill A, Harris M, Sinha AK, Green AC. Prevalence of obesity and metabolic syndrome in Indigenous Australian youths. Obes Rev 2009; 10: 255–261. 62. Carbone ET, Rosal MC, Torres MI, Goins KV, Bermudez OI. Diabetes self-management: perspectives of Latino patients and their health care providers. Patient Educ Couns 2007; 66: 202–210. 63. The Hon Mark Butler MP. New Advisory Council to Boost Disease Prevention Efforts. Canberra. Media release www.health.gov.au; 2011 Jul 18 [3 p]. Available from: http://www.health.gov. au/internet/ministers/publishing.nsf/Content/ 9627AFE36AE57EFCCA2578D10020B581/ $File/mb072.pdf. Accessed 12 February 2012. 64. The Hon Nicola Roxon MP. Telehealth Fees Unveiled [Internet]. Canberra. Media release www.health.gov.au; 2011 Jun 6 [2 p]. Available from: http://www.health.gov.au/internet/ ministers/publishing.nsf/Content/mr-yr11-nrnr116.htm?OpenDocument&yr=2011&mth=06. Accessed 12 February 2012. 65. Australian Government Discussion Paper. Connecting Health Services with The Future: Modernising Medicare By Providing Rebates For Online Consultations. A Discussion Paper from the Australian Government [Internet]. Canberra. Australian Government; 2011 [5 p]. Available from: http://www.health.gov.au/internet/ mbsonline/publishing.nsf/Content/ 256BA3C38B7EEA22CA2577EA006F7C42/ $File/Telehealth per cent20discussion per cent20paper.pdf. Accessed 12 February 2012. 66. The Hon Nicola Roxon MP. More Progress in e-health Consultation and Engagement [Internet]. Canberra: Media Release www.health. gov.au; 2011 Jul 7 [2 p]. Available from: http:// www.health.gov.au/internet/ministers/ publishing.nsf/Content/36DC3163617 C6CCBCA2578C6000773DB/$File/nr138.pdf. Accessed 17 February 2012. 67. National Broadband Network (NBNCo). What the network will look like in your street. [Internet]. Australia. 2010 [updated 2011 Sep]. Available from: http://www.nbnco.com.au/. Accessed 12 February 2012. 68. Australian Government. Department of Broadband, Communication and Digital Economy. #

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

au20 National Digital Economy Strategy [Internet]. Canberra: Department of Broadband, Communication and Digital Economy. 2011 May 31 [68 p]. Available from: http://www.nbn.gov.au/ wpcontent/uploads/2011/05/National_Digital_ Economy_Strategy.pdf. Accessed 12 February 2012. Australian Government. Australian Department of Health and Aging. Draft Concept of Operations: Relating to the Introduction of a personally controlled electronic health record (PCEHR) [Internet]. Canberra: National E-Health Transition Authority (nehta): 2011 Apr [135 p]. Available from: http://www.yourhealth.gov.au/ internet/yourhealth/publishing.nsf/Content/ pcehr. Accessed 12 February 2012. Elliott G, Smith AC, Bensink ME, Brown C, Stewart C, Perry C, Scuffham P. The feasibility of a community-based mobile telehealth screening service for Aboriginal and Torres Strait Islander children in Australia. Telemed J E Health 2010; 16: 950–956. Fonda SJ, Birkmire-Peters D, Bursell SE. Comprehensive Diabetes Management Program (CDMP). In Merrell R, Cooper RA, eds. Proceedings of the Fourth IASTED International Conference on Telehealth and Assistive Technologies Calgary: ACTA Press, 2008. Fonda SJ, Kedziora RJ, Vigersky RA, Bursell SE. Combining iGoogle and personal health records to create a prototype personal health application for diabetes self-management. Telemed J E Health 2010; 16: 480–489. Fonda SJ, Kedziora RJ, Vigersky RA, Bursell SE. Evolution of a web-based, prototype Personal Health Application for diabetes self-management. J Biomed Inform 2010; 43 (5 Suppl): S17–S21. Jordan RE, Lancashire RJ, Adab P. An evaluation of Birmingham Own Health telephone care management service among patients with poorly controlled diabetes. A retrospective comparison with the General Practice Research Database. BMC Public Health 2011; 11: 707. Lim S, Kang SM, Shin H, Lee HJ, Won Yoon J, Yu SH, Kim SY et al. Improved glycemic control without hypoglycemia in elderly diabetic patients using the ubiquitous healthcare service, a new medical information system. Diabetes Care 2011; 34: 308–313. Weinstock RS, Teresi JA, Goland R, Izquierdo R, Palmas W, Eimicke JP, Ebner S et al. Glycemic control and health disparities in older ethnically diverse underserved adults with diabetes: fiveyear results from the Informatics for Diabetes Education and Telemedicine (IDEATel) study. Diabetes Care 2011; 34: 274–279. West SP, Lagua C, Trief PM, Izquierdo R, Weinstock RS. Goal setting using telemedicine in rural underserved older adults with diabetes: experiences from the informatics for diabetes education and telemedicine project. Telemed J E Health 2010; 16: 405–416. Yu CH, Bahniwal R, Laupacis A, Leung E, Orr MS, Straus SE. Systematic review and evaluation of web-accessible tools for management of diabetes and related cardiovascular risk factors by patients and healthcare providers. J Am Med Inform Assoc. 2012; doi:10.1136/amiajnl-2011000307. Welch G, Allen NA, Zagarins SE, Stamp KD, Bursell SE, Kedziora RJ. Comprehensive diabetes

© 2012 The Authors Clinical and Experimental Optometry © 2012 Optometrists Association Australia

Telemedicine and ocular health in diabetes Bursell, Brazionis and Jenkins

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

management program for poorly controlled Hispanic type 2 patients at a community health center. Diabetes Educ 2011; 37: 680–688. Glasgow RE, Hiss RG, Anderson RM, Friedman NM, Hayward RA, Marrero DG, Taylor CB et al. Report of the health care delivery work group: behavioral research related to the establishment of a chronic disease model for diabetes care. Diabetes Care 2001; 24: 124–130. Ersek M, Turner JA, Cain KC, Kemp CA. Results of a randomized controlled trial to examine the efficacy of a chronic pain self-management group for older adults [ISRCTN11899548]. Pain 2008; 138: 29–40. Lorig KR, Holman H. Self-management education: history, definition, outcomes and mechanisms. Ann Behav Med. 2003; 26: 1–7. McGillion MH, Watt-Watson J, Stevens B, Lefort SM, Coyte P, Graham A. Randomized controlled trial of a psychoeducation program for the selfmanagement of chronic cardiac pain. J Pain Symptom Manage 2008; 36: 126–140. Butcher MK, Gilman J, Meszaros JF, Bjorsness D, Madison M, McDowall JM, Oser CS et al. Improving access to quality diabetes education in a rural state: the Montana Quality Diabetes Education Initiative. Diabetes Educ 2006; 32: 963–967. Glazier RH, Bajcar J, Kennie NR, Willson K. A systematic review of interventions to improve diabetes care in socially disadvantaged populations. Diabetes Care 2006; 29: 1675–1688. Lorig KR, Ritter PL, Jacquez A. Outcomes of border health Spanish/English chronic disease self-management programs. Diabetes Educ 2005; 31: 401–409. Lorig KR, Ritter PL, Laurent DD, Plant K. The internet-based arthritis self-management program: a one-year randomized trial for patients with arthritis or fibromyalgia. Arthritia Rheum 2008; 59: 1009–1017. Smith BJ, McElroy HJ, Ruffin RE, Frith PA, Heard AR, Battersby MW, Esterman AJ et al. The effectiveness of coordinated care for people with chronic respiratory disease. Med J Aust 2002; 177: 481–485. Thomas A, Ward E. Peer power: how Dare County, North Carolina, is addressing chronic disease through innovative programming. J Public Health Manage Pract 2006; 12: 462–467. Abramoff MD, Niemeijer M, Russell SR. Automated detection of diabetic retinopathy: barriers to translation into clinical practice. Expert Rev Med Devices 2010; 7: 287–296. Abramoff MD, Niemeijer M, Suttorp-Schulten MS, Viergever MA, Russell SR, van Ginneken B. Evaluation of a system for automatic detection of diabetic retinopathy from color fundus photographs in a large population of patients with diabetes. Diabetes Care 2008; 31: 193– 198. Abramoff MD, Reinhardt JM, Russell SR, Folk JC, Mahajan VB, Niemeijer M, Quellec G. Automated early detection of diabetic retinopathy. Ophthalmology 2010; 117: 1147–1154. Englmeier KH, Schmid K, Hildebrand C, Bichler S, Porta M, Maurino M, Bek T. Early detection of diabetes retinopathy by new algorithms for automatic recognition of vascular changes. Eur J Med Res 2004; 9: 473–478. Fleming AD, Goatman KA, Philip S, Olson JA, Sharp PF. Automatic detection of retinal

95.

96.

97.

98.

99.

100.

101.

102.

103.

anatomy to assist diabetic retinopathy screening. Phys Med Biol 2007; 52: 331–345. Larsen M, Godt J, Larsen N, Lund-Andersen H, Sjolie AK, Agardh E, Kalm H et al. Automated detection of fundus photographic red lesions in diabetic retinopathy. Invest Ophthalmol Vis Sci 2003; 44: 761–766. Niemeijer M, van Ginneken B, Russell SR, Suttorp-Schulten MS, Abramoff MD. Automated detection and differentiation of drusen, exudates, and cotton-wool spots in digital color fundus photographs for diabetic retinopathy diagnosis. Invest Ophthalmol Vis Sci 2007; 48: 2260– 2267. Sanchez CI, Garcia M, Mayo A, Lopez MI, Hornero R. Retinal image analysis based on mixture models to detect hard exudates. Med Image Anal 2009; 13: 650–658. Sivakumar R, Ravindran G, Muthayya M, Lakshminarayanan S, Velmurughendran CU. Diabetic retinopathy analysis. J Biomed Biotechnol 2005; 2005: 20–27. Smith RT, Chan JK, Nagasaki T, Ahmad UF, Barbazetto I, Sparrow J, Figueroa M et al. Automated detection of macular drusen using geometric background leveling and threshold selection. Arch Ophthalmol 2005; 123: 200–206. Staal J, Abramoff MD, Niemeijer M, Viergever MA, van Ginneken B. Ridge-based vessel segmentation in color images of the retina. IEEE Trans Med Imaging 2004; 23: 501–509. Usher D, Dumskyj M, Himaga M, Williamson TH, Nussey S, Boyce J. Automated detection of diabetic retinopathy in digital retinal images: a tool for diabetic retinopathy screening. Diabet Med 2004; 21: 84–90. Chaum E, Karnowski TP, Govindasamy VP, Abdelrahman M, Tobin KW. Automated diagnosis of retinopathy by content-based image retrieval. Retina 2008; 28: 1463–1477. Tobin KW, Abdelrahman M, Chaum E, Govindasamy V, Karnowski TP. A probabilistic framework for content-based diagnosis of retinal disease. Conf Proc IEEE Eng Med Biol Soc 2007; 2007: 6744–6747.

Corresponding author: Professor Sven-Erik Bursell The University of Melbourne Department of Medicine (St Vincent’s) Corner Princes and Regent Streets Fitzroy VIC 3065 AUSTRALIA E-mail: [email protected]

© 2012 The Authors Clinical and Experimental Optometry © 2012 Optometrists Association Australia

Clinical and Experimental Optometry 95.3 May 2012

327