Abstract
Purpose :
In vivo corneal confocal microscopy (CCM) could reveal early corneal microstructural changes in diabetes mellitus before the development of diabetic retinopathy (DR). The aim of our present study was to establish the reliability and repeatability of CCM in young type 1 diabetic patients (T1DM).
Methods :
86 subjects underwent CCM (Heidelberg Retina Tomograph III Rostock Cornea Module; Heidelberg Engineering GmbH, Heidelberg, Germany) examinations, 50 young patients ( mean age of 24.5±7.5 years) with T1DM (34 cases without and 16 cases with DR). The control group comprised 36 age-matched healthy young individuals with a mean age of 23.7±10.6 were included. Using the instrument-based software for cell count, semi-automated image analysis (ACCMetrics; University of Manchester, Manchester, UK) of sub-basal nerve fibers (SBP) were performed independently by two examiners.
Results :
Mostly, there was no significant difference in any of the SBP morphology parameters between examiners. A statistically significant difference was obtained in epithelial and keratocyte cell density in the normal group (p<0.0001, p=0.007) and in epithelial and endothelial cell density in patients without DR (p<0.0001, p=0.004, respectively). For every SBP morphology parameter, the inter-examiner reliability was excellent in the control group, but poor in both diabetic groups. No significant differences were disclosed between 3 consecutive measurements performed by the same examiner except for corneal nerve branch density in patients without DR. Bland-Altman analysis indicated high variability in the measured parameters across patient groups.
Conclusions :
CCM and semi-automated nerve fiber analyzer provided repeatable measurements not only on healthy individuals but also on diabetic patients with and without DR. However, the measurement error showed high variability both in healthy and diabetic subjects. Our results indicate that corneal cell density and SBP morphometry parameters should be measured by the same examiner during a patient follow-up.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.