Retinopathy Telescreening Systems

Number: 0563

Table Of Contents

Applicable CPT / HCPCS / ICD-10 Codes


Scope of Policy

This Clinical Policy Bulletin addresses retinopathy telescreening systems.

  1. Medical Necessity

    Aetna considers retinopathy telescreening systems medically necessary for screening diabetic retinopathy and retinopathy of prematurity as an alternative to retinopathy screening by an ophthalmologist or optometrist.

  2. Experimental and Investigational

    1. Aetna considers artificial intelligence-based systems experimental and investigational for screening of retinopathy of prematurity because the effectiveness of this approach has not been established.
    2. Aetna considers retinopathy telescreening systems experimental and investigational for the following because of insufficient evidence of their clinical value for these indications (not an all-inclusive list):

      • Following the progression of disease in members who are diagnosed with diabetic retinopathy
      • Screening or evaluating retinal conditions other than diabetic retinopathy or retinopathy of prematurity, including, but not limited to macular degeneration/edema.
  3. Related Policies


CPT Codes / HCPCS Codes / ICD-10 Codes

Code Code Description

CPT codes covered if selection criteria are met:

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

CPT codes not covered for indications in the CPB:

Artificial intelligence-based system – no specific code

Other HCPCS codes related to the CPB:

S3000 Diabetic indicator; retinal eye exam, dilated, bilateral

ICD-10 codes covered if selection criteria are met:

E10.10 - E10.29
E10.40 - E10.9
E11.00 - E11.29
E11.40 - E11.9
E13.00 - E13.29
E13.40 - E13.9
Diabetes (except with ophthalmic manifestations)
H35.101 - H35.179 Retinopathy of prematurity
Z13.5 Encounter for screening for eye and ear disorders [screening for retinopathy of prematurity]

ICD-10 codes not covered for indications listed in the CPB (not all-inclusive):

E08.311, E08.3211 - E08.3219, E08.3311 - E08.3319, E08.3411 - E08.3419, E08.3511 - E08.3559, E09.311, E09.3211 - E09.3219, E09.3311 - E09.3319, E09.3411 - E09.3419, E09.3511 - E09.3559, E10.311, E10.3211 - E10.3219, E10.3311 - E10.3319, E10.3411 - E10.3419, E10.3511 - E10.3559, E11.311, E11.3211 - E11.3219, E11.3311 - E11.3319, E11.3411 - E11.3419, E11.3511 - E11.3559, E13.311, E13.3211 - E13.3219, E13.3311 - E13.3319, E13.3411 - E13.3419, E13.3511 - E13.3559 Diabetic retinopathy [not covered for following the progression of disease in members who are diagnosed with diabetic retinopathy]
H35.30 - H35.389 Degenerations of macula and posterior pole
H35.81 Retinal edema
P07.00 - P07.18 Extremely low and other low birth weight newborn [not covered for screening for retinopathy of prematurity]


Diabetic retinopathy is a highly specific vascular complication of both type 1 and type 2 diabetes.  The prevalence of retinopathy is strongly related to the duration of diabetes.  After 20 years of diabetes, nearly all patients with type 1 diabetes and more than 60 % of patients with type 2 diabetes have some degree of retinopathy.  Diabetic retinopathy poses a serious threat to vision.  Overall, diabetic retinopathy is estimated to be the most frequent cause of new cases of blindness among adults aged 20 to 74 years.

Vision loss due to diabetic retinopathy results from several mechanisms.  First, macular edema or capillary non-perfusion may impair central vision.  Second, the new blood vessels of proliferative diabetic retinopathy and contraction of the accompanying fibrous tissue can distort the retina and lead to tractional retinal detachment, producing severe and often irreversible vision loss.  Third, the new blood vessels may bleed, adding the further complication of pre-retinal or vitreous hemorrhage.

One of the main motivations for screening for diabetic retinopathy is the established efficacy of laser photocoagulation surgery in preventing visual loss.  Two large National Institutes of Health sponsored trials, the Diabetic Retinopathy Study and the Early Treatment Diabetic Retinopathy Study, provide the strongest support for the therapeutic benefit of photocoagulation surgery.  Timely laser photocoagulation therapy can prevent loss of vision in a large proportion of patients with severe diabetic retinopathy and/or macular edema.  Since some patients with vision-threatening pathologies may not have symptoms, ongoing evaluation for retinopathy is a valuable and required strategy.

The American Diabetes Association recommends retinopathy screening with yearly retinal examinations beginning at the time of diagnosis of diabetes for all patients age 30 years and older.  For patients under age 30 years, annual retinal examinations are recommended beginning within 3 to 5 years after diagnosis of diabetes once the patient is 10 years old or older.

Diabetic retinopathy telescreening systems involve taking digital pictures of the retina of diabetic patients in the primary care physician's office, and electronically transmitting these pictures to a reading center for evaluation for diabetic retinopathy and macular edema by trained non-physician technicians.  Because diabetic retinopathy telescreening can be performed in conjunction with a primary care physician office visit without referral to an ophthalmologist or optometrist, these systems have the potential to improve compliance with retinopathy screening.  A cost-effectiveness analysis performed by the British National Health Service Centre for Reviews and Dissemination concluded that screening using a digital camera may be more accurate than screening by the general practitioner, and offers an opportunity to reduce costs of diabetic screening, especially as the costs of digital cameras come down.  The UK NHS National Coordinating Centre for Health Technology Assessment (NCCHTA) has initiated a primary research project on the value of digital imaging in diabetic retinopathy.

The Inoveon System of retinopathy screening (iScore, Inoveon Corp., Oklahoma City, OK) involves 7-standard field stereoscopic 30° digital fundus photographs through dilated pupils obtained by a trained photographer located in or near the primary care physician's office, electronic transmission of these digital photographs, examination and grading of these images by non-physician technicians, and rereading of a selected sample of images for assessment of inter- and intra-rater reliability.  In addition, any image sets with questionable pathology or non-typical findings are referred to Inoveon's ophthalmologist medical director for secondary evaluations following initial technician reader evaluation.  If images are not adequate to allow the technician readers to make an assessment of the patient's diabetic retinopathy or macular edema status, Inoveon recommends to the primary care physician that the patient be referred for further evaluation by an ophthalmologist or optometrist.  According to Inoveon Corp., these quality assurance protocols meet or exceed HEDIS specifications for reading centers providing diabetic retinopathy evaluation services.

In a study comparing high-resolution digital stereoscopic fundus photographs (Inoveon System) to plain film stereoscopic fundus photographs (the gold standard), Fransen et al (2002) reported that the digital photographs provided highly accurate diabetic retinopathy referral decisions.  Seven standard field stereoscopic retinal photographs were obtained in 290 adult patients with diabetes by a trained photographer using both a 35-mm plain film camera and a digital camera.  In this double-masked study, each image was independently graded by trained technicians for retinopathy severity (ETDRS severity scale) and for macular edema.  A third technician was used to adjudicate any discrepancies between independent readings.  The primary endpoint was the detection of threshold events requiring referral, which was defined as an ETDRS retinopathy severity level greater than 52, questionable or definite clinically significant macular edema, or ungradable images.  The prevalence of threshold events in the study population was 19.3 %.  The investigators found that the sensitivity of the digital photography system in detecting threshold events, compared to plain film photography, was 98.2 % (confidence interval [CI]: 90.5 % to 100%) and the specificity was 89.7 % (CI: 85.1 % to 93.3 %).  The positive-predictive value was 69.5 % and the negative-predictive value was 99.5 % for this sample.

Rudnisky and colleagues (2002) found high-resolution stereoscopic digital photography comparable to contact lens biomicroscopy in diagnosing clinically significant macular edema.  A total of 120 patients with diabetes underwent clinical examination with contact lens biomicroscopy by a retinal specialist (the gold standard), and on the same day received digital photographs of the macula.  The stereoscopic digital images were evaluated by a single masked grader for the presence or absence of macular edema.  Agreement between digital photographs and contact lens biomicroscopy was 83.6 % for clinically significant macular edema (CSME), 83.6 % for CSME type 1, 96.1 % for CSME type 2, 88.5 % for CSME type 3, 75 % for macular edema, 77.9 % for microaneurysms, 83.7 % for intraretinal hemorrhage, and 73.1 % for hard exudates.  Sensitivity for CSME overall was 90.6 %.  Specificity ranged from 90.0 % for macular edema to 99.0 % for CSME type 2.

The DigiScope Diabetic Retinal Evaluation Service (EyeTel Imaging Corp., Centreville, VA) employs a DigiScope, a specialized digital camera, to obtain high-resolution, wide-field stereoscopic digital images of the retina through dilated pupils.  Trained office personnel use the DigiScope to obtain retinal images.  The DigiScope automatically centers on the pupil, illuminates, focuses, and estimates visual acuity.  The DigiScope images 15 slightly overlapping fields providing a 55 to 60 degree overall view that centered on the macula.  The images are transmitted over phone lines to a central reading center, where the images are evaluated for diabetic retinopathy and macular edema by trained technicians.  The findings are transmitted to the physician and patient.  

An image validation study has demonstrated high correlations between the DigiScope and 7-field stereo color fundus photography as a gold standard (Schiffman et al, 2005).  In a masked prospective study, 111 patients with diabetes (222 eyes) were imaged with the DigiScope and with 7-field stereo color fundus photography.  There was close agreement between the DigiScope and 7-field stereo color fundus photography between "no diabetic retinopathy" and "any diabetic retinopathy" (Kappa statistic 0.97 for the right eye (OD) and 0.94 for the left eye (OS)).  This was reflected in very high sensitivities (0.99 OD, 1.00 OS) and specificities (1.00 OD, 0.92 OS).  As referral on the basis of any retinopathy, no matter how mild, may result in an unnecessarily high number of referrals, the study evaluated a second threshold level of very mild non-proliferative diabetic retinopathy to reduce the number of unnecessary referrals.  Using this threshold, there was substantial agreement based on "microaneurysms or less retinopathy" (which includes no diabetic abnormalities and microaneurysms only) versus retinal hemorrhages or worse retinopathy" (Kappa stastistic 0.78 OD, 0.88 OS), with corresponding sensitivities (0.95 OD, 0.98 OS) and specificities (0.81 OD, 0.87 OS).  The investigators concluded that this image validation study showed that the DigiScope has excellent agreement, sensitivity, and specificity compared with the "gold-standard" 7-field color stereo photography for identifying patients with any or low levels of diabetic retinopathy who should be under the care of an ophthalmologist.  The authors noted, however, that the DigiScope is not designed as a diabetic retinopathy disease management tool or to replace a comprehensive eye examination. 

Recent techniques permit the acquisition of high-quality photographs through undilated pupils and the acquisition of images in digital format.  Although this may eventually permit undilated photographic retinopathy screening, no rigorous studies to date validate the equivalence of these photographs with 7-standard field stereoscopic 30° fundus photography for assessing diabetic retinopathy.  The use of the non-mydriatic camera for follow-up of patients with diabetes in the physician's office might be considered only in situations where dilated eye examinations can not be obtained.

Salcone et al (2010) stated that retinopathy of prematurity (ROP) is a vision-threatening vaso-proliferative condition of premature infants worldwide.  As survival rates of younger and smaller infants improve, more babies are at risk for the development of ROP and blindness.  Meanwhile, fewer ophthalmologists are available for bedside indirect ophthalmoscopy screening examinations.  Remote digital imaging is a promising method with which to identify those infants with treatment-requiring or referral-warranted ROP quickly and accurately, and may help circumvent issues regarding the limited availability of ROP screening providers.  The Retcam imaging system is the most common system for fundus photography, with which high-quality photographs can be obtained by trained non-physician personnel and evaluated by a remote expert.  It has been shown to have high reliability and accuracy in detecting referral-warranted ROP, particularly at later post-menstrual ages.  Additionally, the method is generally well-received by parents and is highly cost-effective.

An UpToDate review on "Retinopathy of prematurity" (Paysse, 2013) states that "screening evaluation consists of a comprehensive eye examination performed by an ophthalmologist with expertise in neonatal disorders".

An American Academy of Ophthalmology Preferred Practice Pattern on Diabetic Retinopathy (AAO, 2014) states: "Some studies have shown that screening programs using digital images taken with or without dilation may enable early detection of diabetic retinopathy along with an appropriate referral. Digital cameras with stereoscopic capabilities are useful for identifying subtle neovascularization and macular edema .... Studies have found a positive association between participating in a photographic screening program and subsequent adherence to receiving recommended comprehensive dilated eye examinations by a clinician. Of course, such screening programs are more relevant when access to ophthalmic care is limited. Screening programs should follow established guidelines. Given the known gap in accessibility of direct ophthalmologic screening, fundus photographic screening programs may help increase the chances that at-risk individuals will be promptly referred for more detailed evaluation and management."   

An UpToDate review on "Age-related macular degeneration: Clinical presentation, etiology, and diagnosis" (Arroyo, 2013) does not mention the use of retinopathy telescreening systems as a management tool.

Telescreening for Retinopathy of Prematurity

In a retrospective analysis, Wang and colleagues (2015) reported the 6-year results of the Stanford University Network for Diagnosis of Retinopathy of Prematurity (SUNDROP) initiative in the context of telemedicine screening initiatives for retinopathy of prematurity (ROP).  Subjects were premature newborns requiring ROP screening at 6 neonatal intensive care units (NICUs) from December 1, 2005, to November 30, 2011.  Infants were evaluated via remote retinal photography by an ROP specialist.  A total of 608 preterm infants meeting ROP examination criteria were screened with the RetCam II/III (Clarity Medical Systems, Pleasanton, CA).  Primary outcomes were treatment-warranted ROP (TW-ROP) and adverse anatomical events.  During the 6 years, 1,216 total eyes were screened during 2,169 examinations, generating 26,970 retinal images, an average of 3.56 examinations and 44.28 images per patient; 22 (3.6 %) of the infants screened met criteria for TW-ROP.  Compared with bedside binocular ophthalmoscopy, remote interpretation of RetCam II/III images had a sensitivity of 100 %, specificity of 99.8 %, positive predicative value (PPV) of 95.5 %, and negative predicative value (NPV) of 100 % for the detection of TW-ROP.  No adverse anatomical outcomes were observed for any enrolled patient.  The authors concluded that the 6-year results for the SUNDROP telemedicine initiative were highly favorable with respect to diagnostic accuracy.  These investigators stated that telemedicine appeared to be a safe, reliable, and cost-effective complement to the efforts of ROP specialists, capable of increasing patient access to screening and focusing the resources of the current ophthalmic community on infants with potentially vision-threatening disease.

On behalf of the AAO, American Academy of Pediatrics (AAP), and American Association of Certified Orthoptists (AACO), Fierson an Capone (2015) noted that ROP remains a significant threat to vision for extremely premature infants despite the availability of therapeutic modalities capable, in most cases, of managing this disorder.  It has been shown in many controlled trials that application of therapies at the appropriate time is essential to successful outcomes in premature infants affected by ROP.  Bedside binocular indirect ophthalmoscopy has been the standard technique for diagnosis and monitoring of ROP in these patients.  However, implementation of routine use of this screening method for at-risk premature infants has presented challenges within the existing care systems, including relative local scarcity of qualified ophthalmologist examiners in some locations and the remote location of some NICUs.  Modern technology, including the development of wide-angle ocular digital fundus photography, coupled with the ability to send digital images electronically to remote locations, has led to the development of telemedicine-based remote digital fundus imaging (RDFI-TM) evaluation techniques.  These techniques have the potential to allow the diagnosis and monitoring of ROP to occur in lieu of the necessity for some repeated on-site examinations in NICUs.  The authors reviewed the currently available literature on RDFI-TM evaluations for ROP and outlined pertinent practical and risk management considerations that should be used when including RDFI-TM in any new or existing ROP care structure.

Wood and co-workers (2016) noted that ROP is a leading cause of childhood blindness.  The incidence of ROP is rising, placing greater demands on the healthcare providers that serve these patients and their families.  Telemedicine remote digital fundus imaging (TM-RDFI) plays a pivotal role in ROP management, and has allowed for the expansion of ROP care into previously underserved areas.  These researchers carried out a broad literature review through the PubMed index with the goal of summarizing the current state of ROP and guidelines for its screening.  Furthermore, all currently used telemedicine remote digital fundus imaging devices were analyzed both via the literature and the companies' websites/brochures.  Finally, the PanoCam LT and PanoCam Pro created by Visunex Medical were analyzed via the company website/brochures.  The authors concluded that the PanoCam LT and PanoCam Pro have recently been approved for use within the U.S. and CE marked for international commercialization in European Union and other countries requiring CE mark.  These wide-field imaging systems have the intended use of ophthalmic imaging of all newborn babies and meet the requirements for ROP screening, thereby serving as competition within the ROP screening market previously dominated by one camera imaging system.

Wongwai and associates (2018) evaluated the diagnostic accuracy of a digital fundus photographic system that consists of taking fundus photographs by a trained technician using a RetCam shuttle and interpreting fundus images by an expert to detect ROP requiring treatment (ROP-RT, which defined as type I ROP according to the Early Treatment for ROP study (ETROP).  A total of 100 infants were examined by an expert ophthalmologist experienced in ROP care using indirect ophthalmoscopy; digital wide-field imaging by a trained technician using a RetCam shuttle and images were sent remotely for interpretation by 2 ophthalmologists experienced in ROP care (Reader A, and Reader B); and local ophthalmologists using indirect ophthalmoscopy.  The diagnostic accuracy consisting of sensitivity, specificity, PPV, NPV, positive likelihood ratio (LR+), and negative likelihood ratio (LR-) were calculated.  Agreement between all examiners and readers were evaluated.  A total of 100 infants (mean gestational age [GA] of 31.1 weeks, mean birth weight of 1,511.1 g) participated in the study; 9 infants were classified as ROP-RT.  Reader A and B had very good agreement in detection of ROP- RT (Kappa 1.00, 95 % CI: 1.00 to 1.00).  For reader A, diagnostic performance parameters (95 % CIs) for detecting ROP-RT were; sensitivity 100.0 % (66.4 to 100.0), specificity 97.8 % (92.1 to 99.7), PPV 81.8 % (48.2 to 97.7), NPV 100.0 % (95.8 to 100.0), LR+ 44.5 (11.3 to 175.2), and LR- 0.1 (0.0 to 0.8).  For reader B these were; sensitivity 100.0 % (66.4 to 100.0), specificity 95.6 % (89.0 to 98.8), PPV 69.2 % (38.6 to 90.9), NPV 100.0 % (95.8 to 100.0), LR+ 22.5 (8.6 to 58.6), LR- 0.1 (0.0 to 0.8).  No adverse events (AEs)were reported.  The authors concluded that diagnosis of ROP-RT from RetCam images taken by trained technicians and evaluated remotely by an expert ophthalmologist had good diagnostic accuracy for screening purposes.

Karkhaneh and colleagues (2019) evaluated sensitivity and specificity of digital retinal image reading in the diagnosis of referral-warranted ROP.  Infants referred to the ROP clinic underwent fundus examination through indirect ophthalmoscopy.  Fundus photographs were acquired using RetCam (shuttle 2; Clarity medical systems, Pleasanton, CA).  Four retinal specialists who were blind to patients' information reviewed the RetCam fundus photographs.  By comparing the results of photographs' readings with that of indirect ophthalmoscopy as the gold standard, the sensitivity and specificity of telescreening was determined.  A total of 147 treatment-naïve patients met the inclusion criteria and were enrolled in the study.  Mean GA was 28.6 ± 2.0 weeks.  Digital retinal imaging had sensitivity of 85 % and specificity of 35 % in detecting referral-warranted ROP in this study; PPV of digital photography was 80 %, and NPV was 43 %.  The authors concluded that considering the large number of ROP patients to be screened, telescreen digital photography could improve the management of patients, prevent significant deleterious visual sequels, and facilitate research.  However, based on the findings of the study, digital imaging cannot be proposed as a substitute for indirect ophthalmoscopy; further study including more patients and graders is needed.

Furthermore, an UpToDate review on "Retinopathy of prematurity: Pathogenesis, epidemiology, classification, and screening" (Coats, 2019) states that "Telemedicine systems can be used to identify infants with potentially severe ROP.  The process involves using wide-angle ocular digital fundus photography to create digital retinal images.  Up to 6 standard images may be taken.  The images are then transmitted to a remote location for interpretation.  Initially, telemedicine was used to provide screening for remote locations without access to an ophthalmologist skilled in ROP screening; however, telemedicine is increasingly used as the primary mode of screening even in locations with access to an ophthalmologist skilled in ROP screening.  When a telemedicine screening approach is used, the AAP/AAO/AAPOS/AACO joint statement suggests that it follow the same schedule as ophthalmoscopic screening (as described above) and that infants at risk undergo indirect ophthalmoscopy by a qualified ophthalmologist at least once before initiating treatment or terminating screening".

Artificial Intelligence-Based Systems for Screening of Retinopathy of Prematurity

In a retrospective analysis of prospectively collected clinical data, Cole et al (2022) examined the performance of a deep learning (DL) algorithm for retinopathy of prematurity (ROP) screening in Nepal and Mongolia.  Clinical information and fundus images were obtained from infants in 2 ROP screening programs in Nepal and Mongolia.  Fundus images were obtained using the Forus 3nethra neo (Forus Health) in Nepal and the RetCam Portable (Natus Medical, Inc.) in Mongolia.  The overall severity of ROP was determined from the medical record using the International Classification of ROP (ICROP).  The presence of plus disease was determined independently in each image using a reference standard diagnosis.  The Imaging and Informatics for ROP (i-ROP) DL algorithm was trained on images from the RetCam to classify plus disease and to assign a vascular severity score (VSS) from 1 through 9.  Main outcome measures were area under the receiver operating characteristic curve (AUROC) and area under the precision-recall curve for the presence of plus disease or type 1 ROP and association between VSS and ICROP disease category.  The prevalence of type 1 ROP was found to be higher in Mongolia (14.0 %) than in Nepal (2.2 %; p < 0.001) in these data sets.  In Mongolia (RetCam images), the AUROC for examination-level plus disease detection was 0.968, and the area under the precision-recall curve was 0.823.  In Nepal (Forus images), these values were 0.999 and 0.993, respectively.  The ROP VSS was associated with ICROP classification in both datasets (p < 0.001).  At the population level, the median VSS was found to be higher in Mongolia (2.7; inter-quartile range [IQR], 1.3 to 5.4]) as compared with Nepal (1.9; IQR, 1.2 to 3.4; p < 0.001).  The authors concluded that these data provided preliminary evidence of the effectiveness of the i-ROP DL algorithm for ROP screening in neonatal populations in Nepal and Mongolia using multiple camera systems and were useful for consideration in future clinical implementation of artificial intelligence (AI)-based ROP screening in low- and middle-income countries (LMIC).  Moreover, these researchers stated that although the road from successful demonstration of AI diagnostic accuracy in an article to clinical implementation at the bedside appeared to be long, these data suggested that “we may be on our way to clinical use of AI-based ROP screening”.

The authors stated that this study had several drawbacks.  First, a number of challenges exist in data collection in the LMIC setting.  Inconsistencies may exist in the recording of clinical data that can lead to some records being unusable for this analysis and in data cleaning between the US-based and internationally-based teams who prepared the datasets.  Despite this, these investigators did not believe that this resulted in systematic bias or otherwise affected the key findings in this trial.  However, a prospective, longitudinal evaluation of the VSS in ROP disease progression is needed to better characterize the clinical usefulness of the VSS and its potential role as an adjunct to clinical diagnosis.  In Mongolia, these researchers employed a secure, web-based ROP database called Research Electronic Data Capture.  In Nepal, data management software called iTeleGEN was implemented.  iTeleGEN enabled the systematic input of data, and further studies are needed to examine its performance.  Second, the precise operating point in actual clinical settings for ROP screening remains to be determined.  This study examined the correlation between the i-ROP DL algorithm and reference standard diagnosis for plus disease as well as the VSS and ICROP category but did not use findings from simulated ROP screenings.  Third, these researchers did not evaluate systematically the impact of image quality on the results, although this will be a key component of clinical implementation.  Despite this, the algorithm demonstrated acceptable performance and might mimic clinical settings more accurately, where it would be difficult to filter every image for quality.

deCampos-Stairiker et al (2023) noted that epidemiological changes in ROP depend on neonatal care, neonatal mortality, and the ability to carefully titrate and monitor oxygen.  In a retrospective, cohort study, these researchers examined if an AI algorithm for evaluating ROP severity in babies can be employed to assess changes in disease epidemiology in babies from South India over a 5-year period.  Subjects were babies (n = 3,093) screened for ROP at neonatal care units (NCUs) across the Aravind Eye Care System (AECS) in South India.  Images and clinical data were collected as part of routine tele-ROP screening at the AECS in India over 2 time-periods: August 2015 to October 2017 and March 2019 to December 2020.  Differences in clinical diagnoses of moderate (type 2) or treatment-requiring (TR) ROP over time were evaluated against birthweight, gestational age at birth, and an AI-derived ROP VSS (derived from retinal fundus images obtained at the initial tele-retinal screening examination).  Main outcome measures were differences in the proportions of type 2 or worse ROP and TR-ROP cases, and differences in VSS between time-periods.  Over time, the proportion [95 % CI] of babies with type 2 or worse ROP and TR-ROP dropped from 60.9 % [53.8 % to 67.7 %] to 17.1%  [14.0 % to 20.5 %] (p < 0.001) and 16.8 % [11.9 % to 22.7 %] to 5.1 % [3.4 % to 7.3 %] (p < 0.001), respectively.  Similarly, the median [IQR] VSS in the population decreased from 2.9 [1.2] to 2.4 [1.8] (p < 0.001).  The authors concluded that these results suggested that AI-based assessment of ROP severity may be a useful epidemiologic tool to examine geographic and temporal changes in ROP incidence, severity, and risk.  In South India, the use of an AI-based measure of ROP severity, when applied at the population level, was strongly suggestive of dramatic improvements in primary prevention of ROP in a short period of time, co-incidental to the implementation of a large ROP tele-screening program.


The above policy is based on the following references:

  1. Aiello LP, Gardner TW, King GL, et al. Diabetic retinopathy. Technical Review. Diabetes Care. 1998;21:143-156. 
  2. American Academy of Ophthalmology (AAO), Retina/Vitreous Panel. Diabetic retinopathy. Preferred Practice Pattern. San Francisco, CA: AAO; November 2014.
  3. American Academy of Ophthalmology (AAO). Diabetic retinopathy. Preferred Practice Pattern. San Francisco, CA: AAO; 2003. 
  4. American Diabetes Association. Position statement: Diabetic retinopathy. Clinical Practice Guidelines 2001. Diabetes Care. 2001; 24 Supp 1:S73-S76.  
  5. Arroyo JG. Age-related macular degeneration: Clinical presentation, etiology, and diagnosis. UpToDate [serial online]. Waltham, MA: UpToDate; reviewed April 2013. 
  6. Bashshur RL, Shannon GW, Smith BR, Woodward MA. The empirical evidence for the telemedicine intervention in diabetes management. Telemed J E Health. 2015;21(5):321-354.
  7. Bek T, Lund-Andersen H, Hansen AB, et al. The prevalence of diabetic retinopathy in patients with screen-detected type 2 diabetes in Denmark: The ADDITION study. Acta Ophthalmol. 2009;87(3):270-274.
  8. Bhaskaranand M, Ramachandra C, Bhat S, et al. Automated diabetic retinopathy screening and monitoring using retinal fundus image analysis. J Diabetes Sci Technol. 2016;10(2):254-261.
  9. Bragge P, Gruen RL, Chau M, et al. Screening for presence or absence of diabetic retinopathy: A meta-analysis. Arch Ophthalmol. 2011;129(4):435-444.
  10. Coats DK. Retinopathy of prematurity: Pathogenesis, epidemiology, classification, and screening. UpToDate [online serial]. Waltham, MA: UpToDate; reviewed March 2019.
  11. Cole E, Valikodath NG, Al-Khaled T, et al. Evaluation of an artificial intelligence system for retinopathy of prematurity screening in Nepal and Mongolia. Ophthalmol Sci. 2022;2(4):100165.
  12. deCampos-Stairiker MA, Coyner AS, Gupta A, et al. Epidemiologic evaluation of retinopathy of prematurity severity in a large telemedicine program in India using artificial intelligence. Ophthalmology. 2023 Apr 6 [Online ahead of print].
  13. Fierson WM, Capone A Jr; American Academy of Pediatrics Section on Ophthalmology; American Academy of Ophthalmology, American Association of Certified Orthoptists. Telemedicine for evaluation of retinopathy of prematurity. Pediatrics. 2015;135(1):e238-e254.
  14. Fransen SR, Leonard-Martin TC, Feuer WJ, et al. Clinical evaluation of patients with diabetic retinopathy. Accuracy of the Inoveon Diabetic Retinopathy-3DT System. Ophthalmology. 2002;109(3):595-601. 
  15. Health Technology Board for Scotland (HTBS). Organisation of services for diabetic retinopathy screening. Health Technology Assessment - consultation report. Glasgow, Scotland: HTBS; April 2002.
  16. Inoveon Corp. Inoveon iScore Diabetic Retinopathy Evaluation Service. Quality assurance program at the Inoveon Evaluation Center, Nashville, Tennessee. Oklahoma City, OK: Inoveon; 2002.
  17. Jani PD, Forbes L, Choudhury A, et al. Evaluation of diabetic retinal screening and factors for ophthalmology referral in a telemedicine network. JAMA Ophthalmol. 2017;135(7):706-714.
  18. Johansen MA, Fossen K, Norum J, et al. The potential of digital monochrome images versus colour slides in telescreening for diabetic retinopathy. J Telemed Telecare. 2008;14(1):27-31.
  19. Joshi GD, Sivaswamy J. DrishtiCare: A telescreening platform for diabetic retinopathy powered with fundus image analysis. J Diabetes Sci Technol. 2011;5(1):23-31.
  20. Karkhaneh R, Ahmadraji A, Riazi Esfahani M, et al. The accuracy of digital imaging in diagnosis of retinopathy of prematurity in Iran: A pilot study. J Ophthalmic Vis Res. 2019;14(1):38-41.
  21. Khandekar R, Senthil T, Nainappan M, Edward DP. Magnitude and determinants of diabetic retinopathy among Indian diabetic patients undergoing telescreening in India. Telemed J E Health. 2022;28(2):176-188.
  22. Kirkizlar E, Serban N, Sisson JA, et al. Evaluation of telemedicine for screening of diabetic retinopathy in the Veterans Health Administration. Ophthalmology. 2013;120(12):2604-2610.
  23. Liesenfeld B. A telemedical approach to the screening of diabetic retinopathy: Digital fundus photography. Diabetes Care. 2000;23(3):345-348. 
  24. Liu L, Geng J, Wu J, et al. Prevalence of ocular fundus pathology with type 2 diabetes in a Chinese urban community as assessed by telescreening. BMJ Open. 2013;3(12):e004146.
  25. Mundy L, Merlin T, Braunack-Mayer A, Hiller JE. Retinal photography and the detection of diabetic retinopathy. Adelaide Health Technology Assessment (AHTA) on behalf of National Horizon Scanning Unit (HealthPACT and MSAC). Adelaide, SA: Department of Public Health, University of Adelaide; 2004.
  26. National Institute for Clinical Excellence (NICE). Management of type II diabetes. Retinopathy - screening and early management - guideline. London, UK: NICE; 2002.
  27. Neubauer AS, Kernt M, Haritoglou C, et al. Nonmydriatic screening for diabetic retinopathy by ultra-widefield scanning laser ophthalmoscopy (Optomap). Graefes Arch Clin Exp Ophthalmol. 2008;246(2):229-235.
  28. NHS Centre for Reviews and Dissemination. Complications of diabetes: Screening for retinopathy; Management of foot ulcers. Effective Health Care. 1999;5(4):1-12. 
  29. Paysse EA. Retinopathy of prematurity. UpToDate [serial online]. Waltham, MA: UpToDate; reviewed April 2013. 
  30. Peng JJ, Huang JN, Lu LN, Zou HD. Research advances in diabetic retinopathy telescreening systems. Zhonghua Yan Ke Za Zhi. 2016;52(11):868-871.
  31. Prasad S, Bannon P, Clearkin LG, et al. Digital fundus imaging: A quality and cost comparison with 35-mm film. Acta Ophthalmol Scand. 1999;77(1):79-82. 
  32. Raman R, Bhojwani DN, Sharma T. How accurate is the diagnosis of diabetic retinopathy on telescreening? The Indian scenario. Rural Remote Health. 2014;14(4):2809.
  33. Ranjan R, Das A, Prema S, et al. Coincidental ocular findings during retinopathy of prematurity telescreening of 9105 neonates. Eur J Ophthalmol. 2022;32(1):563-574.
  34. Rudnisky CJ, Hinz BJ, Tennant MT, et al. High-resolution stereoscopic digital fundus photography versus contact lens biomicroscopy for the detection of clinically significant macular edema. Ophthalmology. 2002;109(2);267-274. 
  35. Schiffman JS, Li HK, Tang RA. Telemedicine enters eye care: Practical experience.  J Ophthalmic Nurs Technol. 1998;17(3):102-106.
  36. Schiffman RM, Jacobsen G, Nussbaum JJ, et al. Comparison of a digital retinal imaging system designed for detection of diabetic retinopathy in the primary care physician’s office to stereo seven-field color fundus photography. Ophthalmol Surg Lasers Imag. 2005;36(1):46-56.
  37. Schneider S, Aldington SJ, Kohner EM, et al. Quality assurance for diabetic retinopathy telescreening. Diabet Med. 2005;22(6):794-802.
  38. Sen P, Wu W-C, Chandra P, et al. Retinopathy of prematurity treatment: Asian perspectives. Eye (Lond). 2020;34(4):632-642.
  39. Sharp PF, Olson J, Strachan F, et al. The value of digital imaging in diabetic retinopathy. Health Technol Assess. 2003;7(30):1-132..
  40. Strul S, Zheng Y, Gangaputra S, et al. Pediatric diabetic retinopathy telescreening. J AAPOS. 2020;24(1):10.e1-10.e5.
  41. Surendran TS, Raman R. Teleophthalmology in diabetic retinopathy. J Diabetes Sci Technol. 2014;8(2):262-266.
  42. Ullah W, Pathan SK, Panchal A, et al. Cost-effectiveness and diagnostic accuracy of telemedicine in macular disease and diabetic retinopathy: A systematic review and meta-analysis. Medicine (Baltimore). 2020;99(25):e20306.
  43. Wang SK, Callaway NF, Wallenstein MB, et al. SUNDROP: Six years of screening for retinopathy of prematurity with telemedicine. Can J Ophthalmol. 2015;50(2):101-106.
  44. Whited JD, Datta SK, Aiello LM, et al. A modeled economic analysis of a digital teleophthalmology system as used by three federal healthcare agencies for detecting proliferative diabetic retinopathy. Telemed J E-Health. 2005;11(6):641-651.
  45. Williams GA, Scott IU, Haller JA, et al. Single-field fundus photography for diabetic retinopathy screening: A report by the American Academy of Ophthalmology. Ophthalmology. 2004;111(5):1055-1062.
  46. Wongwai P, Suwannaraj S, Asawaphureekorn S, et al. Diagnostic accuracy of a digital fundus photographic system for detection of retinopathy of prematurity requiring treatment (ROP-RT). PLoS One. 2018;13(7):e0201544.
  47. Wood EH, Moshfeghi AA, Nudleman ED, Moshfeghi DM. Evaluation of Visunex Medical's PanoCam(TM) LT and PanoCam(TM) Pro wide-field imaging systems for the screening of ROP in newborn infants. Expert Rev Med Devices. 2016;13(8):705-712.
  48. Yogesan K. Telemedicine screening of diabetic retinopathy using a hand-held fundus camera. Telemed J. 2000;6(2):219-223. 
  49. 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(2):89-98.