Comparison of Mydriatic Handheld Retinal Imaging with Early Treatment Diabetic Retinopathy Study (ETDRS) 7-Standard Field Photography for Diabetic Retinopathy (DR) and Diabetic Macular Edema (DME)

Abstract

Purpose : To compare mydriatic handheld retinal imaging with standard ETDRS 7-field color 30-degree fundus photography (ETDRS photos) for assessment of DR/DME

Methods : Following standard imaging protocol, mydriatic retinal images were acquired using handheld retinal cameras [Aurora (AU), Smartscope (SS), RV700 (RV) and InView (NV)] and dilated ETDRS photos (Figure 1). All images were evaluated at a centralized reading center independently by 4 graders (2 certified retinal image/graders, 1 ophthalmologist, 1 retina specialist). All differences were adjudicated by a senior retina specialist. Kappa statistics [simple (K), weighted (K_W)] assessed agreement for DR/ DME. Sensitivity/specificity for any DR, referable DR [(refDR) moderate nonproliferative DR (NPDR) or worse, any DME or ungradable images] and vision threatening DR [(vtDR) severe NPDR or worse, clinically significant DME (CSME) or ungradable images] were calculated.

Results : Images from 177 eyes of 92 patients with diabetes were evaluated. Severity by ETDRS photos: no DR 40.1%, mild NPDR 19.2%, moderate 14.7%, severe in 10.2%, proliferative DR 15.8%; no DME 72.9%, DME 6.8%, ciDME 17.0%; ungradable 3.4%. Ungradable rate for DR/DME for AU: 0%/0.56%; SS: 4.5%/5.7%; RV:4.0%/5.7%; and NV:7.3%/35.6%. Kappa statistics and agreement rates with ETDRS photos for DR/DME are shown in Table 1. Agreement with DR was highest with the AU (K_w=0.79) and lowest with NV (K_w=0.53). DME agreement was similar across all devices (K_w = 0.79-0.83). Agreement for DR severity with ETDRS photos was highest with AU (68.9% exact, 96.6% 1-step). Sensitivity/Specificity for any DR, refDR and vtDR are shown in Table 1. The established standards (0.80 sensitivity,0.95 specificity) were met by AU, SS and RV for any DR and refDR.

Conclusions : Following a standardized protocol, handheld retinal imaging devices have substantial levels of agreement for DR/DME and meet accepted standards for sensitivity and specificity in identifying any DR and refDR. However, the ungradable rate varies greatly, exceeding 35% even with pupil dilation in some devices. None of the handheld devices met the established 95% specificity for vtDR.

This is a 2021 ARVO Annual Meeting abstract.

 

Figure 1 Handheld retinal images and ETDRS photographs

 

Table 1