| Coverage Policy Manual | |
| Policy #: | 2008026 |
| Category: | Medicine |
| Initiated: | September 2008 |
| Last Review: | April 2024 |
|
|
|
|||||||||||
| Digital Imaging for the Detection of Diabetic Retinopathy | |
|
|
|
| Description: |
Digital imaging systems use a digital fundus camera to acquire a series of standard field color images and/or monochromatic images of the retina of each eye. This type of retinopathy screening and risk assessment is proposed as an alternative to conventional dilated fundus examination, particularly in diabetic individuals who are not compliant with the recommended periodic retinopathy screenings. The digital images that are captured may be transmitted via the Internet to a remote center for interpretation by trained readers, storage, and subsequent comparison.
Diabetic retinopathy is the leading cause of blindness among adults aged 20–74 years in the United States. The major risk factors for developing diabetic retinopathy are duration of diabetes and severity of hyperglycemia. After 20 years of disease, almost all patients with type 1 and >60% of patients with type 2 diabetes will have some degree of retinopathy (Garg, 2009). Other important risk factors include hypertension and elevated serum lipid levels.
Diabetic retinopathy progresses, at varying rates, from asymptomatic, mild nonproliferative abnormalities to proliferative diabetic retinopathy (PDR) with new blood vessel growth on the retina and posterior surface of the vitreous. The two most serious complications for vision are diabetic macular edema and proliferative diabetic retinopathy. At its earliest stage (nonproliferative retinopathy), the retina develops microaneurysms, intraretinal hemorrhages, and focal areas of retinal ischemia. With disruption of the blood-retinal barrier, macular retinal vessels become permeable, leading to exudation of serous fluid and lipids into the macula (macular edema). As the disease progresses blood vessels that provide nourishment to the retina are blocked, triggering the growth of new and fragile blood vessels (proliferative retinopathy). The new blood vessels that occur in PDR may fibrose and contract, resulting in tractional retinal detachments with significant vision loss. Severe vision loss with proliferative retinopathy arises from vitreous hemorrhage. Moderate vision loss can also arise from macular edema (fluid accumulating in the center of the macula) during the proliferative or nonproliferative stages of the disease. Although proliferative disease is the main blinding complication of diabetic retinopathy, macular edema is more frequent and is the leading cause of moderate vision loss in people with diabetes.
The value of screening is well established since diabetic retinopathy has few visual or ocular symptoms until vision loss develops. With early detection, diabetic retinopathy can be treated with modalities that can decrease the risk of severe vision loss. Tight glycemic and blood pressure control is the first line of treatment to control diabetic retinopathy, followed by laser photocoagulation for patients whose retinopathy is approaching the high-risk stage. Although laser photocoagulation is effective at slowing the progression of retinopathy and reducing visual loss, it results in collateral damage to the retina and does not restore lost vision. Focal macular edema (characterized by leakage from discrete microaneurysms on fluorescein angiography) may be treated with focal laser photocoagulation, while diffuse macular edema (characterized by generalized macular edema on fluorescein angiography) may be treated with grid laser photocoagulation. Corticosteroids may reduce vascular permeability and inhibit vascular endothelial growth factor (VEGF) production, but are associated with serious adverse effects including cataracts and glaucoma with damage to the optic nerve. Corticosteroids also can worsen diabetes control. VEGF inhibitors (e.g., ranibizumab, bevacizumab, and pegaptanib), which reduce permeability and block the pathway leading to new blood vessel formation (angiogenesis), are being evaluated for the treatment of diabetic macular edema and proliferative diabetic retinopathy.
Because treatments are aimed primarily at preventing vision loss, and retinopathy can be asymptomatic, it is important to detect disease and begin treatment early in the process. Annual dilated, indirect ophthalmoscopy coupled with biomicroscopy or 7-standard field stereoscopic 30° fundus photography have been considered to be the screening techniques of choice. Because these techniques require a dedicated visit to a competent eye care professional, typically an ophthalmologist, there is underutilization of this screening recommendation by at-risk members. The under-use has resulted in the exploration of remote retinal imaging, using film or digital photography, as an alternative to direct ophthalmic examination of the retina.
A number of photographic methods have been evaluated that allow images of the retina to be captured and then interpreted by expert readers who may not be located conveniently to the patient. One approach is mydriatic standard field 35-mm stereoscopic color fundus photographs. Digital fundus photography has also been evaluated as an alternative to conventional film photography. Retinal imaging can be performed using digital retinal photographs with (mydriatic) or without (nonmydriatic) dilating the pupil. Digital imaging has the advantage of easier acquisition, transmission, and storage. In addition, the potential for digital images of the retina to be acquired in a primary care setting and evaluated by trained readers in a remote location with retinal specialist consultation exists.
Several digital camera and transmission systems have been cleared for marketing by the U.S. Food and Drug Administration (FDA) through the 510(k) process and are currently available. FDA product codes: HKI and NFJ. In 2018, the FDA gave De Novo clearance for the automated retinal analysis system (IDx-DR) that uses artificial intelligence (DEN180001). IDx-DR is indicated "for use by health care providers to automatically detect more than mild diabetic retinopathy in adults diagnosed with diabetes who have not been previously diagnosed with diabetic retinopathy. IDx-DR is indicated for use with the Topcon NW400."
Examples of Digital Camera and Transmission Systems Cleared by FDA for Retinal Telescreening include the following:
Automated Analysis Systems include:
Coding
Effective in January 2011, there are specific CPT codes for this testing:
92227: Remote imaging for detection of retinal disease (e.g., retinopathy in a patient with diabetes) with analysis and report under physician supervision, unilateral or bilateral
92228: Remote imaging for monitoring and management of active retinal disease (e.g., diabetic retinopathy) with physician review, interpretation and report, unilateral or bilateral
Effective January 1, 2021, CPT codes 92227 and 92228 were updated, and CPT code 92229 was added:
92227: Imaging of retina for detection or monitoring of disease; with remote clinical staff review and report, unilateral or bilateral
92228: Imaging of retina for detection or monitoring of disease; with remote physician or other qualified health care professional interpretation and report, unilateral or bilateral
92229: Imaging of retina for detection or monitoring of disease; point of care automated analysis and report, unilateral or bilateral
|
|
|
|
|
Policy/ Coverage: |
This policy does not apply to 92250 (fundus photography, with interpretation and report).
Effective March 2021
Meets Primary Coverage Criteria Or Is Covered For Contracts Without Primary Coverage Criteria
The use of digital imaging with mydriasis and manual grading of images meets member benefit certificate primary coverage criteria that there be scientific evidence of effectiveness in improving health outcomes only for screening for diabetic retinopathy in patients with a diagnosis of diabetes mellitus with no previous diagnosis of diabetic retinopathy.
Digital retinal imaging with image interpretation by artificial intelligence software that is approved by the U.S. Food and Drug Administration (e.g., IDX-DR, EyeArt) for the screening of diabetic retinopathy in patients with a diagnosis of diabetes mellitus with no previous diagnosis of diabetic retinopathy meets member benefit certificate primary coverage criteria that there be scientific evidence of effectiveness in improving health outcomes.
Does Not Meet Primary Coverage Criteria Or Is Investigational For Contracts Without Primary Coverage Criteria
Digital imaging of the retina without mydriasis or digital imaging with mydriasis for any other use not described above, including the monitoring and management of disease in individuals diagnosed with diabetic retinopathy does not meet primary coverage criteria that there be scientific evidence of effectiveness.
For contracts without primary coverage criteria, digital imaging of the retina without mydriasis or digital imaging with mydriasis for any other use not described above, including the monitoring and management of disease in individuals diagnosed with diabetic retinopathy is considered investigational. Investigational services are specific exclusions in most member benefit certificates of coverage.
Effective April 2020-March 2021
Meets Primary Coverage Criteria Or Is Covered For Contracts Without Primary Coverage Criteria
The use of digital imaging with mydriasis meets member benefit certificate primary coverage criteria that there be scientific evidence of effectiveness in improving health outcomes only for screening for diabetic retinopathy in patients with a diagnosis of diabetes mellitus with no previous diagnosis of diabetic retinopathy.
Does Not Meet Primary Coverage Criteria Or Is Investigational For Contracts Without Primary Coverage Criteria
Digital imaging of the retina without mydriasis or digital imaging with mydriasis for any other use, including the monitoring and management of disease in individuals diagnosed with diabetic retinopathy does not meet Primary Coverage Criteria that there be scientific evidence of effectiveness.
For contracts without Primary Coverage Criteria, digital imaging of the retina without mydriasis or digital imaging with mydriasis for any other use, including the monitoring and management of disease in individuals diagnosed with diabetic retinopathy is considered investigational. Investigational services are specific exclusions in most member benefit certificates of coverage.
Digital retinal imaging with automated image interpretation for the detection of diabetic retinopathy does not meet primary coverage criteria that there be scientific evidence of effectiveness.
For contracts without primary coverage criteria, digital retinal imaging with automated image interpretation for the detection of diabetic retinopathy is considered investigational. Investigational services are specific contract exclusions in most member benefit certificates of coverage
Effective Prior to April 2020
Meets Primary Coverage Criteria Or Is Covered For Contracts Without Primary Coverage Criteria
The use of digital imaging with mydriasis meets member benefit certificate primary coverage criteria that there be scientific evidence of effectiveness in improving health outcomes only for screening for diabetic retinopathy in patients with a diagnosis of diabetes mellitus (ICD-9 250.0-250.93) with no previous diagnosis of diabetic retinopathy.
Does Not Meet Primary Coverage Criteria Or Is Investigational For Contracts Without Primary Coverage Criteria
Digital imaging of the retina without mydriasis or digital imaging with mydriasis for any other use, including the monitoring and management of disease in individuals diagnosed with diabetic retinopathy does not meet Primary Coverage Criteria that there be scientific evidence of effectiveness. The Criteria exclude coverage of interventions if there is a lack of scientific evidence regarding the intervention, or if the available scientific evidence is in conflict or the subject of continuing debate.
For contracts without Primary Coverage Criteria, digital imaging of the retina without mydriasis or digital imaging with mydriasis for any other use, including the monitoring and management of disease in individuals diagnosed with diabetic retinopathy is considered investigational. Investigational services are specific exclusions in most member benefit certificates of coverage.
Effective Prior to February 2018
The use of digital imaging with mydriasis meets member benefit certificate primary coverage criteria that there be scientific evidence of effectiveness in improving health outcomes only for screening for diabetic retinopathy in patients with a diagnosis of diabetes mellitus (ICD-9 250.0-250.93) with no previous diagnosis of diabetic retinopathy.
Digital imaging of the retina without mydriasis or digital imaging with mydriasis for any other use, including the monitoring and management of disease in individuals diagnosed with diabetic retinopathy does not meet Primary Coverage Criteria that there be scientific evidence of effectiveness. The Criteria exclude coverage of interventions if there is a lack of scientific evidence regarding the intervention, or if the available scientific evidence is in conflict or the subject of continuing debate.
For contracts without Primary Coverage Criteria, digital imaging of the retina without mydriasis or digital imaging with mydriasis for any other use, including the monitoring and management of disease in individuals diagnosed with diabetic retinopathy is considered investigational. Investigational services are specific exclusions in most member benefit certificates of coverage.
Effective prior to September 2012
The use of digital imaging with mydriasis meets member benefit certificate primary coverage criteria that there be scientific evidence of effectiveness in improving health outcomes only for screening for diabetic retinopathy in patients with a diagnosis of diabetes mellitus (ICD-9 250.0-250.9) with no previous diagnosis of diabetic retinopathy.
Digital imaging of the retina without mydriasis or for any other use does not meet Primary Coverage Criteria that there be scientific evidence of effectiveness. The Criteria exclude coverage of interventions if there is a lack of scientific evidence regarding the intervention, or if the available scientific evidence is in conflict or the subject of continuing debate.
For contracts without Primary Coverage Criteria, digital imaging of the retina without mydriasis, or digital imaging with mydriasis for any condition other than the screening of those with diabetes mellitus is considered investigational. Investigational services are specific exclusions in most member benefit certificates of coverage.
|
|
|
|
| Rationale: |
Diabetic retinopathy is the most frequent cause of new cases of blindness among adults aged 20–74 years. The likelihood of retinal changes increases with length of time the condition has been existent. After 20 years of disease, almost all patients with type 1 and >60% of patients with type 2 diabetes will manifest microvascular changes characteristic of the disease. (Fong, 2004)
Retinopathy progresses, at varying rates, from asymptomatic, mild nonproliferative abnormalities to proliferative diabetic retinopathy (PDR) with new blood vessel growth on the retina and posterior surface of the vitreous. Resultant vision loss may be due to any of several mechanisms. Central or acute vision may be impaired from edema that develops from increased vascular permeability or from ischemia from capillary nonperfusion. The new blood vessels that occur in PDR may stimulate the development of fibrous tissue resulting in bleeding and traction retinal detachments. (ETDRS, 1991)
The value of screening for diabetic retinopathy is well established. Diabetic retinopathy has few visual or ocular symptoms until vision loss develops. Laser photocoagulation is effective at retarding the progression of the changes but uncommonly is able to restore lost vision. Because treatments are aimed at preventing vision loss, and retinopathy can be asymptomatic, it is important to detect disease and begin treatment early in the process. The benefit of early treatment of diabetic retinopathy was established in the large Early Treatment Diabetic Retinopathy Study (ETDRS) supported by the National Eye Institute (NEI). (2003)
Annual dilated, indirect ophthalmoscopy coupled with biomicroscopy or 7-standard field stereoscopic 30°fundus photography have been considered to be the screening techniques of choice. (Aiello, 1998) Because these techniques require a dedicated visit to a competent eye care professional, typically an ophthalmologist, there is underutilization of this screening recommendation by at-risk members. The under-use rate is estimated to be 30% or higher, (EDTRS, 1998) which has resulted in the exploration of retinal imaging, using film or digital photography, as an alternative to direct ophthalmic examination of the retina.
A number of photographic methods have been evaluated that allow images of the retina to be captured and then interpreted by expert readers who may not be located conveniently to the patient. This local acquisition/remote interpretation technique was used to consistently detect and evaluate the retinal changes of participants in the ETDRS. The ETDRS used 7 mydriatic standard field 35-mm stereoscopic color fundus photographs evaluated by a single reading center. Digital fundus photography has been evaluated as an alternative to conventional film photography. (1998) Digital imaging has the advantage of easier acquisition, transmission, and storage. In addition, the potential for digital images of the retina to be acquired in a primary care setting and evaluated by trained readers in a remote location with retinal specialist consultation exists.
A number of studies have reported on the agreement regarding the presence and stage of retinopathy based on ophthalmoscopy versus photography or standard film versus digital imaging. The studies generally found a high level of agreement between retinal examination and imaging. Several studies suggested that retinal imaging through a dilated pupil was equivalent or superior to ophthalmic examination regarding the detection of diabetic retinal changes. Moss et al (1985) reported on an overall agreement of 85.7% when comparing retinopathy detection by ophthalmoscopy performed by skilled examiners to 7-standard field stereoscopic 30°fundus photography evaluated by trained graders. A study by Kinyoun, 1992, found fair-to-good agreement between ophthalmoscopy and evaluating of 7-standard field stereoscopic 30°fundus photography by the examining ophthalmologist as well as by trained readers. Analysis of the discordance suggested that conventional ophthalmoscopy could miss up to 50% of microaneurysms, some of the earliest changes of diabetic retinopathy. Delori et al (1977) reported more accurate visualization and documentation of the structures of the ocular fundus when using monochromatic illumination (red-free green light) as compared to the white light used to obtain color photographs.
The efficacy of digital image acquisition, as compared to film-based acquisition, has been reported by several investigators. (Tennant, 2001; Liesenfeld, 2000) Fransen et al (2002) published the results of a comparison of standard evaluations using film to the same fields captured and transmitted as digital images. In the study of 290 adult diabetic patients, the sensitivity of digital compared to film was 98.2% and the specificity was 98.7%. Statistical analysis identified that the evaluation of film and digital images provided substantially equivalent results. When comparing high-resolution stereoscopic digital fundus photography to contact lens biomicroscopy, Rudnisky et al (2002) found a high level of agreement regarding the detection of clinically significant macular edema in diabetic patients.
In addition to the examination technique and the comparison of different photographic techniques, the results of dilated versus nonmydriatic fundus photography has been studied. (Heaven, 1993; Peters, 1993; Scanlon, 2003) compared mydriatic and nonmydriatic photo screening programs using dilated slit lamp biomicroscopy as the reference standard. In the study of 3,611 patients, the sensitivity of mydriatic digital photography was 87.8%, the specificity was 86.1%, and the technical failure rate was 3.7%. Photography through an undilated pupil was found to provide a sensitivity of 86.0%, a specificity of 76.6%, and a technical failure rate of 19.7%. The authors concluded that while dilated digital photography is an effective method of screening for diabetic retinopathy, nonmydriatic photography has an unacceptable failure rate and low specificity.
Chun et al (2007) compared a single 45 degree, nonmydriatic digital color photograph with dilated ophthalmoscopy in 231 eyes of 120 patients and found the level of agreement only moderate for both diabetic retinopathy and cystoid macular edema. They concluded a single 45 degrees, nonmydriatic, digital image was not reliable as a sole modality to screen for diabetic retinopathy.
Deb-Joardar et al (2005) reported on 150 patients, 300 eyes, who had five field (45 degree) digital retinal imaging and mosaic construction through dark-adapted pupils and then after a single application of tropicamide 1%. Of 300 eyes pharmacological mydriasis improved image quality with an increase in the number of eyes with 5 good images to 160 from 7; the number of eyes having retinopathy detected with certainty went from 153 to 252 and graded with certainty went from 173 to 277. There were no adverse effects noted.
There is published medical literature adequate to conclude that digital imaging systems are safe and effective alternatives to the gold standards of dilated indirect ophthalmoscopy coupled with biomicroscopy or stereoscopic fundus photography. Additional advantages of digital imaging systems include short examination time, ability to perform without mydriasis, and the ability to perform the test in the primary care physician setting.
Studies continue to report that digital imaging systems are an acceptable alternative to a dilated fundus examination for the evaluation of diabetic retinopathy, and are adaptable for use in the primary care physicians’ office. (Wilson, 2005; Hansen, 2004)
2011 Update
A 2011 meta-analysis evaluated variations in qualifications of photographers and mydriatic status (Bragge, 2011). Twenty studies were included that evaluated the accuracy of a diabetic retinopathy screening method that used photography- or examination-based retinopathy screening compared with a standard of either 7-field mydriatic photography or dilated fundal examination. Studies with film or digital cameras were included in the systematic review. Studies of automated analysis techniques and technologies were excluded because they were not considered current standard practice. For meta-analysis, 40 assessments of screening methods were collapsed into 6 categories: nonmydriatic camera, nonspecialist photographer (n=5); mydriatic camera, nonspecialist photographer (n=8); nonmydriatic camera, specialist photographer (n=4); mydriatic camera, specialist photographer (n=3); direct examination (n=8); method mixed or not reported (n=12). Sensitivity and specificity were assessed for the presence or absence of diabetic retinopathy in comparison with the reference standard. Variations in mydriatic status alone did not significantly influence sensitivity (odds ratio, [OR] 0.89) or specificity (OR, 0.94). Variations in medical qualifications of photographers did not significantly influence sensitivity (OR, 1.25), but the specificity of detection of any diabetic retinopathy was significantly higher for screening methods that used a photographer with specialist medical or eye qualifications. When photographs were taken by a specialist, the odds of a negative screening test when diabetic retinopathy was not evident with the reference standard were 3.86 times that when photographs were taken by nonspecialists. This was largely due to the effect of specialists or nonspecialists in photographs taken without mydriasis (OR 5.65). The lower specificity with nonspecialist photographers may lead to increased referrals to an eye specialist for further examination in some patients without diabetic retinopathy. This finding may be biased, since 6 of 7 assessments in the specialist category were derived from a single study. Interpretation is further limited by the inclusion of both standard film and digital imaging in the meta-analysis.
In 2010 the American Diabetes Association (ADA) updated their position statement on standards of medical care in diabetes. Included in the guidelines are specific recommendations for initial and subsequent examinations to screen for retinopathy (ADA, 2010). The ADA states that examinations can be done with retinal photographs (with or without dilation of the pupil) read by experienced experts. In-person exams are still necessary when the photos are unacceptable and for follow-up of abnormalities detected.
In 2011, the American Telemedicine Association (ATA) published guidelines for clinical, technical, and operational performance standards for diabetic retinopathy screening (ATA, 2011). Recommendations from the ATA are based on reviews of current evidence, medical literature and clinical practice. The ATA states that ETDRS thirty-degree, stereo seven-standard field, color, 35 mm slides are an accepted standard for evaluating diabetic retinopathy. Although no standard criteria have been widely accepted as performance measurements of digital imagery used for diabetic retinopathy evaluation, current clinical trials sponsored by the National Eye Institute have transitioned to digital images for diabetic retinopathy assessment. Telehealth programs for diabetic retinopathy should demonstrate an ability to compare favorably with ETDRS film or digital photography as reflected in kappa values for agreement of diagnosis, false positive and false negative readings, positive predictive value, negative predictive value, sensitivity and specificity of diagnosing levels of retinopathy and macular edema. Inability to obtain or read images should be considered a positive finding and patients with unobtainable or unreadable images should be promptly re-imaged or referred for evaluation by an eye care specialist.
There was no new literature identified that would prompt a change in the coverage statement.
2012 Update
A search of the MEDLINE database through August 2012 did not reveal any new information that would prompt a change in the coverage statement.
A number of automated scoring systems are being evaluated for diabetic retinopathy screening. A 2011 publication examined the accuracy of one such approach, which used a computer-aided diagnosis (CAD) system to diagnose diabetic retinopathy using a publicly available dataset of 1,200 digital color fundus photographs (Sanchez, 2011). The reference standard was based on 2 diagnoses provided with the dataset. At a specificity of 50%, the automated system had a sensitivity of 92.2% to detect diabetic retinopathy, which was similar to the results of 2 expert reviewers (sensitivity of 94.5% and 91.2% and specificity of 50%). Fifty-one abnormal images were wrongly classified as normal. Research is continuing to improve the system’s performance.
Oliveira et al. assessed the accuracy of another automated screening system (RetmarkerSR) in a study of non-mydriatic images from 5,386 patients in a diabetic retinopathy screening program (Oliveira, 2011). Automated analysis classified 47.5% as having no disease and 52.5% as having disease. When compared with an experienced ophthalmologist grader who graded 8.7% with referable retinopathy, the sensitivity was 96.1% and specificity was 51.7%. A 2-step approach in which patients marked as diseased on the first screen had a second screening visit improved specificity to 63.2% with no loss of sensitivity. The sample in this study was biased, as it did not include another 9.54% of images that a grader had identified as being of poor quality. The omission of these cases may have led to a falsely high estimate of accuracy.
2014 Update
A literature search conducted through August 2014 did not reveal any new information that would prompt a change in the coverage statement.
2016 Update
A literature search conducted through August 2016 did not reveal any new information that would prompt a change in the coverage statement. The key identified literature is summarized below.
The efficacy of diabetic retinopathy detection with digital image acquisition, compared with film-based acquisition, has been reported by several investigators.
In 2015, Shi and colleagues reported a systematic review and meta-analysis of studies that compared telemedicine (digital image acquisition) with 7-field fundus photography for the detection of diabetic retinopathy or diabetic macular edema (Shi, 2015). Twenty studies with a total of 1960 patients were included in the qualitative analysis; however, 4 of these reported on the same patient population, so 17 studies were included in the meta-analysis. Studies varied in the specific digital photography techniques used in terms of the number of fields evaluated the use of stereoscopic versus monoscopic and mydriatic versus non mydriatic techniques. In pooled analysis, the sensitivity of digital imaging with telemedicine ophthalmologic evaluation for various diabetic retinopathy states (presence/absence of diabetic retinopathy, mild, moderate, or severe nonproliferative diabetic retinopathy, high- and low-risk proliferative diabetic retinopathy, diabetic macular edema, and clinically significant macular edema) was greater than 70%, except for the detection of severe nonproliferative diabetic retinopathy (sensitivity 53%, 95% confidence interval [CI] 45% to 62%). In pooled analysis, the specificity of digital imaging for various diabetic retinopathy states was greater than 90%, except for the detection of mild nonproliferative diabetic retinopathy (specificity 89%, 95% CI 88% to 91%). Summary receiver operating characteristic (ROC) curves showed an area under the curve (AUC) of greater than 0.9 for the detection of diabetic retinopathy and diabetic macular edema, across a range of severity.
Examples of individual studies which report on the diagnostic accuracy of digital image acquisition include those by Liesenfeld and colleagues (Liesenfeld, 2000) and Tennant and colleagues (2001), which report high correlation between diabetic retinopathy diagnoses made by slit-lamp biomcroscopy performed by an ophthalmologist or by 7-field 35-mm photography, respectively.
One study was identified that evaluated the effectiveness of a telemedicine screening program for diabetic retinopathy, compared with traditional surveillance with an eye care professional, in the setting of a randomized clinical trial (RCT) (Mansberger, 2015). The study randomized 567 adult patients with diabetes to a telemedicine program (n=296) or traditional surveillance (n=271). After 2 years of enrollment, those randomized to the traditional surveillance program were offered the opportunity to cross over to telemedicine screening. The telemedicine photography protocol involved the capture of 6 un-dilated 45°fundus photographs of each eye, with grading of the retinal images by 2 investigators into 5 categories of retinopathy and for the presence of macular edema. At 0-6 months of follow-up, those randomized to the telemedicine program were more likely to undergo retinopathy screening compared with those randomized to traditional surveillance: 94.6% vs 43.9% (risk difference 50.7%, 95% CI 46.6% to 54.8%, P<0.001). There was also a significant difference in screening rates at 6-18 months of follow-up: 53.0% in the telemedicine group vs 33.2% in the traditional screening group (risk difference 19.8%, 95% CI 16.5 to 23.1%, P<0.001). Beyond 18 months, when telemedicine was offered to all participants, there were no significant differences in screening rates between the 2 groups. Throughout follow-up, most subjects (greater than 90%) had a diabetic retinopathy stage within ±1 stage of their baseline stage.
Rasmussen and colleagues compared the agreement of diabetic retinopathy screening results obtained with ETDRS 7-field fundus photography with those obtained from single-image mydriatic widefield photography, nonmydriatic widefield photography, and mydriatic steered photography among 95 diabetic patients (Rasmussen, 23015). Exact agreement between the nonmydriatic widefield photography and the 7-field fundus photography occurred in 76.3% of cases (κ 0.71, 95% CI 0.69 to 0.72). However, agreement within 1 level of retinopathy occurred in 98.9% of cases (κ 0.98, 95% CI 0.95 to 0.99).
In a large retrospective study including 15,015 individuals with diabetes, Walton and colleagues compared manual interpretation of nonmydriatic fundus images with the Intelligent Retinal Imaging System (IRIS), an automated computer algorithm-based interpretation system, in the detection of sight-threatening diabetic eye disease (STDED; severe nonproliferative diabetic retinopathy or proliferative diabetic retinopathy (Walton, 2015); of 18,025 patients with fundus photographs obtained as part of a county screening program, 15,015 (83.3%) had photographs available for IRIS analysis. Compared with centralized manual interpretation, in screening population, the IRIS algorithm had the following sensitivity, specificity, and positive and negative predictive values for STDED: 66.4% (95% CI 62.8 to 69.9%), 72.8% (95% CI 72.0 to 72.5%), 10.8% (95% CI 9.6 to 11.9%), and 97.8% (95% CI 96.8 to 98.6%), respectively.
The evidence for digital retinal photography with optometrist or ophthalmologist image interpretation for individuals who have diabetes without known diabetic retinopathy includes retrospective studies reporting on the accuracy of digital screening compared with standard methods, systematic reviews of these studies, and 1 randomized controlled trial (RCT). Relevant outcomes include test accuracy, test validity, change in disease status, and functional outcomes. A number of studies have reported on the agreement between direct ophthalmoscopy and photography and between standard film and digital imaging in terms of the presence and stage of retinopathy. The studies generally found a high level of agreement between retinal examination and imaging. There is limited direct evidence related to visual outcomes for patients evaluated with a strategy of retinal telescreening. However, the evidence from the large Early Treatment Diabetic Retinopathy Study (ETDRS) that early retinopathy treatment improves outcomes coupled with studies showing high concordance between the screening methods used in ETDRS and one RCT demonstrating higher uptake of screening with a telescreening strategy, a strong chain of evidence can be made that telescreening is associated with improved health outcomes. Digital imaging systems have the additional advantages of short examination time and the ability to perform the test in the primary care physician setting. For individuals who cannot or would not be able to access an eye professional at the recommended screening intervals, the use of telescreening has low risk and is very likely to increase the likelihood of retinopathy detection. The evidence is sufficient to determine qualitatively that the technology is likely to improve the net health outcome.
The evidence for digital retinal photography with automated image interpretation for individuals who have diabetes without known diabetic retinopathy includes retrospective studies reporting on the accuracy of automated scoring of digital images compared with standard methods. Relevant outcomes include test accuracy, test validity, change in disease status, and functional outcomes. The available studies tend to report high sensitivity with moderate specificity, although there is variability across studies. These scoring systems have potential to improve screening in the primary care setting. However, given the variability in test characteristics across different systems, there is uncertainty about the accuracy of automated scoring systems in practice. The evidence is insufficient to determine the effects of the technology on health outcomes.
Practice Guidelines and Position Statements
American Diabetes Association
In 2016 the American Diabetes Association (ADA) updated their position statement on standards of medical care in diabetes (previous updates ADM 2010 and Fong,2004) (ADM, 2016). Included in the guidelines are specific recommendations for initial and subsequent examinations to screen for retinopathy (see Policy Guidelines section). These guidelines make the following statements related to telescreening for retinopathy:
“Retinal photography, with remote reading by experts, has great potential to provide screening services in areas where qualified eye care professionals are not readily available. High-quality fundus photographs can detect most clinically significant diabetic retinopathy. Interpretation of the images should be performed by a trained eye care provider…In-person exams are still necessary when the retinal photos are unacceptable and for follow-up if abnormalities are detected. Retinal photos are not a substitute for a comprehensive eye exam, which should be performed at least initially and at intervals thereafter as recommended by an eye care professional. Results of eye examinations should be documented and transmitted to the referring health care professional.”
2017 Update
A literature search conducted through August 2017 did not reveal any new information that would prompt a change in the coverage statement. The key identified literature is summarized below.
The Iowa Detection Program (IDP), an automated screening system, uses standardized algorithms to detect various retinal findings. This system was evaluated with a publicly available sample of digital color photographs from 1748 eyes (874 patients with diabetes) that were at risk for diabetic retinopathy (Abramoff, 2013). The photographs were taken in primary care diabetic retinopathy clinics from 3 hospitals in France and then graded by 3 masked retinal specialists. The prevalence of referable diabetic retinopathy (more than mild nonproliferative retinopathy and/or macular edema) was 21.7% (95% CI, 19.0% to 24.5%). The diagnostic characteristics of the IDP, compared with expert consensus standard, are summarized in Table 1. The area under the receiver operating curve was 0.937.
In 2016, the same study group reporting on the algorithm evaluated a deep learning algorithm add-on to the IDP algorithm, using the same dataset as in their 2013 study (Abramoff, 2016). Also in 2016 and 2017, Tufail and colleagues reported on the screening performance of 4 automated retinal image analysis systems in a retrospective, observational study, which included 20,258 patients seen for diabetes eye screening, run by the National Health Service from 2012 to 2013 (Tufail, Tufail, 2016; Tufail, 2017).The manual images were graded by a team of 18 optometrists and nonoptometrists who had undergone pre-study training and evaluation. The automated scoring systems identified included EyeArt (Eyenuk, Woodland Hills, CA), Retmarker (Retmarker, Coimbra, Portugal), iGradingM (Medalytix Group, now EMIS UK, Leeds, England), and IDx-DR (IDx, Iowa City, IA). However, the iGradingM was determined to be unable to process disc-centered images, and IDx withdrew from the study, so details on their test performance are not discussed here further. The overall prevalence of referable diabetic retinopathy (defined as ungradable images, maculopathy, and pre-proliferative and proliferative retinopathy) was 2767 (13.7%) of 20,212. Compared with manual grading, referable diabetic retinopathy using the EyeArt and Retmarker systems was associated with likelihood ratios of 1.375 (95% CI, 1.354 to 1.4) and 1.63 (95% CI, 1.59 to 1.66), respectively.
2019 Update
Annual policy review completed with a literature search using the MEDLINE database through February 2019. No new literature was identified that would prompt a change in the coverage statement.
2020 Update
Annual policy review completed with a literature search using the MEDLINE database through December 2019. No new literature was identified that would prompt a change in the coverage statement. The key identified literature is summarized below.
Abramoff et al in 2018 published the pivotal study of the IDx-DRAI image analysis (Abramoff, 2018). This was a non-inferiority trial that compared the AI analysis system with expert mydriatic photography and centralized reading of images. Nine hundred patients with diabetes and no history of diabetic retinopathy were enrolled at primary care centers. The primary care staff received four hours of training in image capture and use of the system. The system includes an image quality algorithm, which recommended pupil dilation in 23.6% of patients when 3 attempts at nonmydriatic image capture had failed. Compared to expert mydriatic photography and centralized image assessment, the AI system had sensitivity of 87.2%, specificity of 90.7%, PPV of 74.9% and negative predictive value of 95.7%.
2020 Update (April 2020)
The pivotal study of the IDx-DR AI image analysis system (DEN180001) was published by Abramoff et al (Abramoff, 2018). The reference standard was expert mydriatic photography with centralized reading of images. Performance thresholds for the FDA application were set at 85.0% for sensitivity and 82.5% for specificity. Nine hundred patients with diabetes and no history of diabetic retinopathy were enrolled at primary care centers. The study was enriched with patients who had elevated hemoglobin A1c in order to increase the likelihood of enrolling patients with more serious diabetic retinopathy. The primary care staff received 4 hours of training in image capture and use of the system. The system includes an image quality algorithm, which recommended pupil dilation in 23.6% of patients when 3 attempts at nonmydriatic image capture had failed. Compared to expert mydriatic photography and centralized image assessment, the AI system had sensitivity of 87.2%, specificity of 90.7%, PPV of 74.9% and negative predictive value of 95.7%. Enrichment corrected sensitivity and specificity calculated similar diagnostic performance if the study population had not been enriched with subjects with higher hemoglobin A1c levels.
2021 Update
Annual policy review completed with a literature search using the MEDLINE database through January 2021. No new literature was identified that would prompt a change in the coverage statement.
2021 Update
Policy review completed with a literature search using the MEDLINE database through February 2021. The key identified literature is summarized below.
The pivotal study for the EyeArt 2.1.0 artificial intelligence imaging system (NCT03112005) was reported in the summary of the 510(k)application to the U.S. Food and Drug Administration (FDA, 2021). In addition to 235 participants who were sequentially enrolled, an enriched cohort of 420 participants was studied. Participants were seen in either a primary care setting or an ophthalmology setting. Initial 2-field non-mydriatic images were automatically analyzed by EyeArt, which notified the operator if the image was not gradable in order to retake images. Imageability on the first attempt ranged from 83.5% to 94.2%. This was then followed with a reference standard of mydriatic 4-wide field images that were graded at a centralized reading facility. For the non-enriched cohort, prevalence of more than mild diabetic retinopathy was present in 12.2% of patients seen in primary care and10.5% of patients seen by ophthalmologists. Sensitivity for more than mild diabetic retinopathy was 100% among primary care providers and 92.5% by ophthalmologists. Specificity was 88.5% among primary care providers and 85.7% for ophthalmologists. For the enriched cohort of 335 patients seen in primary care, disease prevalence was 15.5%, with sensitivity of 92.9%, and specificity of85.6%. For the enriched cohort seen in ophthalmology practices, disease prevalence was 19.4% with sensitivity of 96.6% and specificity of 85.2% to detect more than mild diabetic retinopathy. Heydon et al reported a prospective independent evaluation of the EyeArt v2.1.0 analysis system in over 30,000 patients from the English Diabetic Eye Screening Programme (Heydon, 2020). The purpose of the study was to assess the utility of the automated analysis system as a screening tool when used in conjunction with human graders. The cameras used and the graders differed between the 3 sites. Images that had been previously scored by human graders were submitted for analysis by EyeArt and classified as referable(positive n=15,091) or non-referable (negative n=15,314). Images that were ungradable by EyeArt were considered referable for further evaluation. Overall, sensitivity and specificity were 95.7% and 54.0%, respectively. EyeArt classified for referral (positive) all cases that had been graded as moderate-to-severe retinopathy by human graders (sensitivity of 100%) but would not have referred 78 (10.6%) of the 739 images that were considered ungradable by the human graders. The number of false positives was high, but it was estimated that when used as a primary screening tool the software could reduce the workload of first level human graders by half.
Lee et al evaluated diagnostic accuracy to detect referable retinopathy with 7 different artificial intelligence algorithms in a sample of over 26,000 patients from 2 Veteran Affairs Health Systems (Lee, 2021). The same camera (Topcon TRC-NW8) was used for all images, but the centers differed on whether the images were mydriatic or non-mydriatic. Over 16% of non-mydriatic images were ungradable compared to 2.5% of mydriatic images. For the analysis, 5 manufacturers (OpthAI, AirDoc, Eyenuk, RetinaAI Health,Retmarker) provided their locked software preloaded on a workstation; the software was identified only by letters A to G. All artificial intelligence algorithms were used clinically across the world, and 1 (EyeArt by Ayenuk) was cleared by the FDA for marketing at the time of the study. Across the 7 algorithms, sensitivity ranged from 50.98% to 85.90%, and specificity ranged from 60.42% to 83.69%,indicating that each marketed software needs to be evaluated separately. Only one of the algorithms had diagnostic performance equal to the human teleretinal graders.
Use of the EyeArt image analysis software was also tested in a study of 69 patients from a retina clinic who were screened using a smartphone-based camera (RetinaScope) by non-ophthalmic personnel (Kim 2021). Compared to the gold standard evaluation by a retina specialist, automated interpretation of images had a sensitivity of 87.0% and specificity of 78.6%; grader 1 had a sensitivity of 96.3% and specificity of 42.9%; grader 2 had a sensitivity of 92.5% and specificity of 50.0%. Further study in a larger, more diverse, sample is needed.
2022 Update
Annual policy review completed with a literature search using the MEDLINE database through February 2022. No new literature was identified that would prompt a change in the coverage statement. The key identified literature is summarized below.
Full results from the EyeArt 2.1.0 pivotal study were published in 2021 and confirmed the accuracy of the system to detect both more-than-mild diabetic retinopathy (sensitivity 95.5%; 95% CI, 92.4% to 98.5%; specificity85.0%; 95%CI, 82.6% to 87.4%) and vision-threatening diabetic retinopathy (sensitivity 95.1%; 95% CI, 90.1% to 100%; specificity89.0%; 95% CI, 87.0% to 91.1%) without dilation (Ipp, 2021).
2023 Update
Annual policy review completed with a literature search using the MEDLINE database through February 2023. No new literature was identified that would prompt a change in the coverage statement.
2024 Update
Annual policy review completed with a literature search using the MEDLINE database through February 2024. No new literature was identified that would prompt a change in the coverage statement.
2024 Update
Annual policy review completed with a literature search using the MEDLINE database through March 2024. No new literature was identified that would prompt a change in the coverage statement.
|
|
|
|
| CPT/HCPCS: | |
|
|
|
| References: |
Abramoff MD, Folk JC, Han DP, et al.(2013) Automated analysis of retinal images for detection of referable diabetic retinopathy. JAMA Ophthalmol. Mar 2013;131(3):351-357. PMID 23494039 Abramoff MD, Lou Y, Erginay A, et al.(2016) Improved automated detection of diabetic retinopathy on a publicly available dataset through integration of deep learning. Invest Ophthalmol Vis Sci. Oct 01 2016;57(13):5200-5206. PMID 27701631 Abramoff, MD, Lavin, PT, Birch, M, Shah, N, et al.(2018) Pivotal trial of an autonomous AI-based diagnostic system for detection of diabetic retinopathy in primary. npj Digital Medicine (2018) 1:39; doi:10.1038/s41746-018-0040-6. Aiello LP, Gardner TW, et al.(1998) Diabetic retinopathy. Diabetes Care. 1998; 21(1):143-56. American Diabetes Association.(2016) 9. Microvascular Complications and Foot Care. Diabetes Care. Jan 2016;39 Suppl 1:S72-80. PMID 26696685 American Telemedicine Association.(2011) Telehealth practice recommendations for diabetic retinopathy. 2011 Available online at: http://www.americantelemed.org/files/public/standards/DiabeticRetinopathy_withCOVER.pdf. Last accessed May, 2011. Bawankar P, Shanbhag N, Smitha KS, et al.(2017) Sensitivity and specificity of automated analysis of single-field non-mydriatic fundus photographs by Bosch DR Algorithm-Comparison with mydriatic fundus photography (ETDRS) for screening in undiagnosed diabetic retinopathy. PLoS One. Dec 27 2017;12(12):e0189854. PMID 29281690 Bragge P, Gruen RL, Chau M et al.(2011) Screening for Presence or Absence of Diabetic Retinopathy: A Meta-analysis. Arch Ophthalmol 2011; 129(4):435-44. Chun DW, Bauer RM, et al.(2007) Evaluation of digital fundus images as a diagnostic method for surveillance of diabetic retinopathy. Mil Med. 2007; 172(4):405-10. Deb-Joardar N, Germain N, et al.(2005) Screening for diabetic retinopathy by ophthalmologists and endocrinologists with pupillary dilation and a nonmydriatic digital camera. Am J Ophthal 2005; 140(5):814-21. Early Treatment Diabetic Retinopathy Study Research Group.(1991) Fundus photographic risk factors for progression of diabetic retinopathy. Ophthalmology. 1991; 98(5 suppl):823-33. Early Treatment Diabetic Retinopathy Study Research Group.(1991) Grading diabetic retinopathy from stereoscopic color fundus photographs: an extension of the modified Airlie House classifications. Ophthalmology. 1991; 98(suppl 5):786-806. Fong DS, Aiello L, et al.(2004) Retinopathy in diabetes. Diabetes Care. 2004; 27(suppl 1):S84-7. Fransen SR, Leonard-Martin TC, et al.(2002) Clinical evaluation of patients with diabetic retinopathyL accuracy of the Inoveon diabetic retinopathy 3DT system. Ophthalmology. 2002; 109(3):595-601. Hansen AB, Hartvig NV, et al.(2004) Diabetic retinopathy screening using digital non-mydriatic fundus photography and automated image analysis. Acta Ophthal Scand. 2004; 82(6):666-72. Heaven CJ, Cansfield J, Shaw KM.(1993) The quality of photographs produced by the non-mydriatic fundus camera in a screening programme for diatbetic retinopathy. Eye. 1993l 7(pt 6):787-90. Heydon P, Egan C, Bolter L, et al.(2020) Prospective evaluation of an artificial intelligence-enabled algorithm for automated diabetic retinopathy screening of 30 000 patients. Br J Ophthalmol. Jun 30 2020. PMID 32606081 Horton MB, Brady CJ, Cavallerano J, et al.(2020) Practice Guidelines for Ocular Telehealth-Diabetic Retinopathy, Third Edition. Telemed J EHealth. Apr 2020; 26(4): 495-543. PMID 32209018 Ipp E, Liljenquist D, Bode B, et al.(2021) Pivotal Evaluation of an Artificial Intelligence System for Autonomous Detection of Referrableand Vision-Threatening Diabetic Retinopathy. JAMA Netw Open. Nov 01 2021; 4(11): e2134254. PMID 34779843 Kim TN, Aaberg MT, Li P, et al.(2021) Comparison of automated and expert human grading of diabetic retinopathy using smartphone-based retinal photography. Eye (Lond). Jan 2021; 35(1): 334-342. PMID 32341536 Kinyoun JL, Martin DC, et al.(1992) Ophthalmoscopy versus fundus photographs for detecting and grading diabetic retinopathy. Invest Ophthalmol Vis Sci. 1992; 33(6):1888-93. Lee AY, Yanagihara RT, Lee CS, et al.(2021) Multicenter, Head-to-Head, Real-World Validation Study of Seven Automated Artificial Intelligence Diabetic Retinopathy Screening Systems. Diabetes Care. Jan 05 2021. PMID 33402366 Liesenfeld B, Kohner E, et al.(2000) A telemedical approach to the screeing of diabetic retinopathy: digital fundus photography. Diabetes Care. 2000; 23(3):345-8. Mansberger SL, Sheppler C, Barker G, et al.(2015) Long-term comparative effectiveness of telemedicine in providing diabetic retinopathy screening examinations: a randomized clinical trial. JAMA Ophthalmol. May 2015;133(5):518-525. PMID 25741666 Moss SE, Klein R, et al.(1985) Comparison between ophthalmoscopy and fundus photography in determining severity of diabetic retinopathy. Ophthamology. 1985; 92(1):62-7. Oliveira CM, Cristovao LM, Ribeiro ML et al.(2011) Improved automated screening of diabetic retinopathy. Ophthalmologica 2011; 226(4):191-7. Oliveira CM, Cristovao LM, Ribeiro ML, et al.(2011) Improved automated screening of diabetic retinopathy. Ophthalmologica. 2011;226(4):191-197. PMID 21865671 Peters AL, Davidson MB, Ziel FH.(1993) Cost-effective screening for diabetic retinopathy using a nonmydriatic retinal camera in a prepaid health-care setting. Diabetes Care. 1993; 16(8):1193-5. Rasmussen ML, Broe R, Frydkjaer-Olsen U, et al.(2015) Comparison between Early Treatment Diabetic Retinopathy Study 7-field retinal photos and non-mydriatic, mydriatic and mydriatic steered widefield scanning laser ophthalmoscopy for assessment of diabetic retinopathy. J Diabetes Complications. Jan-Feb 2015;29(1):99-104. PMID 25240716 Rudnisky CJ, Hinz BJ, et al.(2002) High-resolution stereoscopic digital fundus photography versus contact lens biomicroscopy for the detection of clinically significant macular edema. Ophthalmology. 2002; 109(2):267-74. Sanchez CI, Niemeijer M, Dumitrescu AV et al.(2011) Evaluation of a computer-aided diagnosis system for diabetic retinopathy screening on public data. Invest Ophthalmol Vis Sci 2011; 52(7):4866-71. Sanchez CI, Niemeijer M, Dumitrescu AV, et al.(2011) Evaluation of a computer-aided diagnosis system for diabetic retinopathy screening on public data. Invest Ophthalmol Vis Sci. Jun 2011;52(7):4866-4871. PMID 21527381 Scanlon PH, Malhotra R, et al.(2003) The effectiveness of screening for diabetic retinopathy by digital imaging photography and technician ophthalmoscopy. Diabet Med. 2003; 20(6):467-74. Shi L, Wu H, Dong J, et al.(2015) Telemedicine for detecting diabetic retinopathy: a systematic review and metaanalysis. Br J Ophthalmol. Jun 2015;99(6):823-831. PMID 25563767 Tennant MT, Greve MD, et al.(2001) Identification of diabetic retinopathy by stereoscopic digital imaging via teleophthalmology: a comparison to slide film. Can J Ophthalmol. 2002; 36(4):187-96. Tufail A, Kapetanakis VV, Salas-Vega S, et al.(2016) An observational study to assess if automated diabetic retinopathy image assessment software can replace one or more steps of manual imaging grading and to determine their cost-effectiveness. Health Technol Assess. Dec 2016;20(92):1-72. PMID 27981917 Tufail A, Rudisill C, Egan C, et al.(2017) Automated diabetic retinopathy image assessment software: diagnostic accuracy and cost-effectiveness compared with human graders. Ophthalmology. Mar 2017;124(3):343-351. PMID 28024825 U.S. Food and Drug Administration.(2005) Center for Drug Evaluation and Research, Application number 21-737, Medical Review. 2005; 2005; http://www.accessdata.fda.gov/drugsatfda_docs/nda/2005/021737s000_MedR.pdf. Accessed January 27, 2020. http://www.accessdata.fda.gov/drugsatfda_docs/nda/2005/021737s000_MedR.pdf. Accessed January 27, 2020. U.S. Food and Drug Administration.(2021) EyeArt 510(k) Summary (K200667). August 3, 2020. Available at: https://www.accessdata.fda.gov/cdrh_docs/pdf20/K200667.pdf. Accessed February 10, 2021. Walton OBt, Garoon RB, Weng CY, et al.(2015) Evaluation of automated teleretinal screening program for diabetic retinopathy. JAMA Ophthalmol. Dec 17 2015:1-6. PMID 26720694 Wilson C, Horton M, et al.(2005) Additon 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; 26(2):318-22. |
|
|
|
|
Group specific policy will supersede this policy when applicable. This policy does not apply to the Wal-Mart Associates Group Health Plan participants or to the Tyson Group Health Plan participants.
CPT Codes Copyright © 2024 American Medical Association. |
|