Screening for diabetic retinopathy in diabetic patients with a mydriasis-free, full-field flicker electroretinogram recording device
Yunkao Zeng,Dan Cao,Dawei Yang,Xuenan Zhuang,Honghua Yu,Yunyan Hu,Yan Zhang,Cheng Yang,Miao He,Liang Zhang
DOI: https://doi.org/10.1007/s10633-019-09734-2
2019-11-12
Documenta Ophthalmologica
Abstract:<h3 class="Heading">Purpose</h3><p class="Para">To investigate the accuracy of the RETeval full-field flicker ERG in the screening of diabetic retinopathy (DR) and vision-threatening diabetic retinopathy (VTDR) and to determine a suitable range of DR diagnostic reference for patients with type 2 diabetes mellitus (T2DM).</p><h3 class="Heading">Methods</h3><p class="Para">This was a cross-sectional study involving 172 subjects with T2DM, including 71 subjects without clinically detectable DR (NDR), 25 subjects with mild non-proliferative diabetic retinopathy (NPDR), 24 subjects with moderate NPDR, 27 subjects with severe NPDR and 25 subjects with proliferative diabetic retinopathy (PDR). All the subjects underwent a full-field flicker ERG using the RETeval device (DR assessment protocol), which is a mydriasis-free, full-field electroretinogram (ERG) recording system. The performance of the DR assessment protocol in detecting the DR (including mild NPDR, moderate NPDR, severe NPDR and PDR) and VTDR was analyzed with the receiver operating characteristic (ROC) curve.</p><h3 class="Heading">Results</h3><p class="Para">For the detection of DR (mild NPDR, moderate NPDR, severe NPDR, PDR), the area under the ROC curve was 0.867 (<em class="EmphasisTypeItalic">p</em> < 0.001, 95% CI 0.814–0.920), and the best cutoff value for DR was determined to be 20.75, with a sensitivity of 80.2% and specificity of 81.7%. Meanwhile, for the detection of VTDR, the area under the ROC curve was 0.965 (<em class="EmphasisTypeItalic">p</em> < 0.001, 95% CI 0.941–0.989), and the best cutoff value was set to 23.05, with a sensitivity of 94.6% and a specificity of 88.8%.</p><h3 class="Heading">Conclusion</h3><p class="Para">The DR assessment protocol in RETeval device was effective in screening for DR (mild NPDR, moderate NPDR, severe NPDR, PDR) and VTDR in patients with diabetes. It could be helpful in referring and managing patients with T2DM in primary healthcare setting. However, caution should be taken that optimal cutoff value of DR assessment protocol may vary in different ethnic populations.</p>
ophthalmology