A randomised trial of non-mydriatic ultra-wide field retinal imaging versus usual care to screen for diabetic eye disease: rationale and protocol for the Clearsight trial

Validation of diagnostic accuracy of retinal image grading by trained non-ophthalmologist grader for detecting diabetic retinopathy and diabetic macular edema

PURPOSE

To validate the fundus image grading results by a trained grader (Non-ophthalmologist) and an ophthalmologist grader for detecting diabetic retinopathy (DR) and diabetic macular oedema (DMO) against fundus examination by a retina specialist (gold standard).

METHODS

A prospective diagnostic accuracy study was conducted using 2002 non-mydriatic colour fundus images from 1001 patients aged ≥40 years. Using the Aravind Diabetic Retinopathy Evaluation Software (ADRES) images were graded by both a trained non-ophthalmologist grader (grader-1) and an ophthalmologist (grader-2). Sensitivity, specificity, positive predictive value and negative predictive value were calculated for grader-1 and grader-2 against the grading results by an independent retina specialist who performed dilated fundus examination for every study participant.

RESULTS

Out of 1001 patients included, 42% were women and the mean ± (SD) age was 55.8 (8.39) years. For moderate or worse DR, the sensitivity and specificity for grading by grader-1 with respect to the gold standard was 66.9% and 91.0% respectively and the same for the ophthalmologist was 83.6% and 80.3% respectively. For referable DMO, grader-1 and grader-2 had a sensitivity of 74.6% and 85.6% respectively and a specificity of 83.7% and 79.8% respectively.

CONCLUSIONS

Our results demonstrate good level of accuracy for the fundus image grading performed by a trained non-ophthalmologist which was comparable with the grading by an ophthalmologist. Engaging trained non-ophthalmologists potentially can enhance the efficiency of DR diagnosis using fundus images. Further study with multiple non-ophthalmologist graders is needed to verify the results and strategies to improve agreement for DMO diagnosis are needed.

Effectiveness of Nonmydriatic Ultra-Widefield Retinal Imaging to Screen for Diabetic Eye Disease: A Randomized Controlled Trial (Clearsight)

OBJECTIVE

Suboptimal diabetic eye disease screening is a major cause of preventable vision loss. Screening barriers include mydriasis and the need for dedicated screening appointments. The Clearsight trial assessed whether nonmydriatic ultra-widefield (NM UWF) screening on the day of a diabetes clinic visit improved detection of clinically important eye disease versus usual screening.

RESEARCH DESIGN AND METHODS

This single-center, randomized, parallel-group controlled trial was conducted at St. Joseph's Health Care, London, Ontario, Canada. Adults with diabetes due for screening were randomized to same-day, on-site screening (NM UWF imaging) on the day of a scheduled diabetes clinic visit or usual screening (encouraged to arrange optometrist screening). The primary outcome was detection of actionable eye disease (AED), defined as the need for an ophthalmology referral or increased ocular surveillance. The primary analysis (modified intention-to-screen) compared the proportions of AED between groups within 1 year of enrollment.

RESULTS

Of 740 participants randomized between 7 March 2016 and 17 April 2019, 335 on-site screening and 323 usual screening participants met criteria for the primary analysis. More AED was detected in the on-site screening group than in the usual screening group (50 of 335 [14.9%] vs. 22 of 323 [6.8%]; adjusted odds ratio 2.51; 95% CI 1.49-4.36). The number needed to screen by on-site screening in order to detect 1 additional patient with AED was 13 (95% CI 8-29).

CONCLUSIONS

Same-day, on-site screening by NM UWF imaging increased the detection of clinically important diabetic eye disease versus usual screening. Integration of NM UWF imaging into routine diabetes clinic visits improved screening adherence and has the potential to prevent vision loss.

Inter-observer agreement in grading severity of diabetic retinopathy in wide-field fundus photographs

OBJECTIVE

To examine the inter-observer agreement between two retina specialists in grading diabetic retinopathy (DR) severity in ultra-wide-field fundus photographs.

METHODS

Two hundred and seventy patients with diabetes, who visited the vitreoretinal specialty at a tertiary eye care hospital, with or without DR underwent comprehensive ophthalmic examination, dilated retinal exam and Optos ultra-wide-field (UWF) retinal photography. Optos images were graded for DR severity based on the International Clinical Diabetic Retinopathy Disease Severity Scale by two retina specialists with same number of years of experience, masked to the clinical details of the participants.

RESULTS

The two graders showed agreement in 229/270 images (84.8%) and disagreement in 41/270 images (15.2%). The unweighted kappa for agreement between graders was k = 0.715, SE = 0.037 and the weighted kappa was k = 0.838, SE = 0.022. No DR was identified in 170/270 (62.9%) patients, mild NPDR in 15/270 (5.6%) patients, moderate NPDR in 35/270 (12.9%) patients, severe NPDR in 4/270 (1.48%) patient and PDR in 5/270 (1.85%) patients by both graders. Disagreement was neither related to the learning curve of graders nor with the patient's age (p = 0.574), gender (p = 0.169), duration of diabetes (0.660) or the lens being phakic or pseudophakic (p = 0.171) on logistic regression.

CONCLUSIONS

The impact of disagreement noted between observers in grading DR on UWF fundus photographs should be considered when utilizing UWF system in clinical studies.

Effectiveness of Teleretinal Imaging-Based Hospital Referral Compared With Universal Referral in Identifying Diabetic Retinopathy: A Cluster Randomized Clinical Trial

Question

Does screening for diabetic retinopathy by teleretinal imaging in physician offices in India lead to higher adherence to eye hospital referral and a greater yield of diabetic retinopathy cases compared with a strategy of referral of all eligible patients with diabetes?

Findings

In a cluster randomized clinical trial of 801 patients with diabetes, proportionately more patients in the teleretinal group attended the hospital eye examination and had confirmed diabetic retinopathy compared with the control group.

Meaning

The results suggest that, in the Indian setting, teleretinal screening is an effective approach for identifying diabetic retinopathy.

Importance

Studies in high-income countries provide limited evidence from randomized clinical trials on the benefits of teleretinal screening to identify diabetic retinopathy (DR).

Objective

To evaluate the effectiveness of teleretinal-screening hospital referral (TR) compared with universal hospital referral (UR) in people with diabetes.

Design, Setting, and Participants

A cluster randomized clinical trial of 8 diabetes clinics within 10 km from Aravind Eye Hospital (AEH), Madurai, India, was conducted. Participants included 801 patients older than 50 years. The study was conducted from May 21, 2014, to February 7, 2015; data analysis was performed from March 12 to June 16, 2015.

Interventions

In the TR cohort, nonmydriatic, 3-field, 45° retinal images were remotely graded by a retinal specialist and patients with DR, probable DR, or ungradable images were referred to AEH for a retinal examination. In the UR cohort, all patients were referred for a retinal examination at AEH.

Main Outcomes and Measures

Hospital-diagnosed DR.

Results

Of the 801 participants, 401 were women (50.1%) (mean [SD] age, 60.0 [7.3] years); mean diabetes duration was 8.6 (6.6) years. In the TR cohort, 96 of 398 patients (24.1%) who underwent teleretinal imaging were referred with probable DR (53 [13.3%]) or nongradable images (43 [10.8%]). Hospital attendance at AEH was proportionately higher with TR (54 of 96 referred [56.3%]) compared with UR (150 of 400 referred [37.5%]). The intention-to-treat analysis based on all patients eligible for referral in each arm showed that proportionately more patients with TR (36 of 96 [37.5]%) were diagnosed with DR compared with UR (50 of 400 [12.5%]) (unadjusted risk ratio [RR], 3.00; 95% CI, 2.01-4.48). These results were little changed by inclusion of covariates (RR, 2.72; 95% CI, 1.90-3.91). The RR was lower in the per-protocol analysis based on all patients who adhered to referral (covariate-adjusted RR, 1.75; 95% CI, 1.12-2.74). Diagnoses of DR were predominantly mild or moderate nonproliferative DR (36 in TR and 43 in UR). In the UR arm, there were 4 cases of severe nonproliferative DR and 2 cases of proliferative DR. Age (RR, 0.98; 95% CI, 0.95-0.99), female sex (RR, 0.79; 95% CI, 0.64-0.98), and hypertension diagnosis (RR, 0.81; 95% CI, 0.68-0.95) were factors associated with lower attendance. Those with higher secondary educational level or more were twice as likely to attend (RR, 2.00; 95% CI, 1.32-3.03).

Conclusions and Relevance

The proportionate yield of DR cases was higher in the TR arm, confirming the potential benefit, at least in the setting of eye hospitals in India, of a targeted referral approach using teleretinal screening to identify patients with DR.

Trial Registration

ClinicalTrials.gov identifier: NCT02085681