Purpose: To investigate the repeatability of macular and peripapillary retinal nerve fiber layer (RNFL) thickness measurements made using swept-source optical coherence tomography (SS-OCT) and automated segmentation. Measurements were made in non-diabetic controls and in patients with diabetic retinopathy (DR) with or without diabetic macular edema (DME).
Materials and methods: A total of 131 eyes of 131 participants were included. Fifty-one eyes with DR had no DME (DME[-]), 45 eyes with DR had DME (DME[+]), and 35 eyes were healthy. Measurements of RNFL and full retinal thickness were simultaneously obtained with SS-OCT in the peripapillary area and in the nine Early Treatment Diabetic Retinopathy Study (ETDRS) subfields using the wide three-dimensional mode. All measurements were made twice on the same day by a single examiner to test intra-observer repeatability. Intraclass correlation coefficients (ICCs) and coefficients of repeatability were examined to evaluate repeatability.
Results: Average macular and temporal peripapillary RNFL thickness values were greater in the DME[+] group (36.4 ± 13.2 and 83.8 ± 19.4 µm, respectively) than in the control (27.4 ± 3.5 and 73.5 ± 11.4 µm, respectively) and DME[-] (27.9 ± 3.4 µm and 70.3 ± 11.3 µm, respectively) groups (both P < 0.001). The ICCs of average macular (control: 0.982, DME[-]: 0.913, and DME[+]: 0.970) and peripapillary (control: 0.972, DME[-]: 0.973, and DME[+]: 0.958) RNFL thickness measurements indicated good repeatability in all three study groups.
Conclusions: Although the ICCs of average RNFL thickness measurements were relatively lower in eyes with DR than in healthy controls, the intra-observer repeatability of SS-OCT RNFL and full retinal thickness measurements is sufficiently reliable for them to be clinically useful.
Keywords: Diabetic macular edema; diabetic retinopathy; repeatability; retinal nerve fiber layer; swept-source optical coherence tomography.