Reliability and reproducibility of retinal oxygen saturation measurements using a predefined peri-papillary annulus.

PURPOSE: To determine the reliability and reproducibility of the Vesselmap Oximetry Module for arteriolar and venular oxygen saturation (SO2 ) of the same retinal area, specified by a peripapillary annulus, in healthy subjects. METHODS: Fundus oximetry images were obtained, using a standardized protocol by a single observer, from the right eye of 20 healthy individuals. Age range was 19-45 years old, and images were analysed using the oximetry module of the Vesselmap System (Imedos, UG, Germany). Intra-observer reliability (assessment of two measurements of SO2 values performed 5 days apart); interobserver reliability (assessment of SO2 performed by two different trained observers); and intrasubject reproducibility (comparison of SO2 measurements of two different images of the same subject and retinal area, taken 10 min apart) were assessed. The standard deviations between the repeated measurements (SDr), together with the intraclass correlation coefficient (ICC), of these three parameters were calculated. RESULTS: The SDr for intra-observer reliability was 0.56% and 0.55% for arteriolar and venular SO2 , respectively. The results were similar for intrasubject reproducibility (0.69% and 0.79% for arteriolar and venular SO2, respectively); interobserver reliability, however, was higher (SDr 1.22% and 1.01% for arteriolar and venular SO2 , respectively). The ICC values for intra-observer reliability were 0.99 for both arteriolar and venular SO2 . The results were similar for both interobserver reliability (0.94 for arteriolar SO2 and 0.96 for venular SO2 ) and intrasubject reproducibility (0.98 for both arteriolar and venular SO2 ). CONCLUSION: Retinal oxygen saturation values taken using the oximetry module of the Vesselmap System are highly reliable and reproducible, provided the image quality is standardized, the same measurement area is analysed in each image, and the number of observers analysing the images is kept to a minimum.

Duke Scholars

Author Tien Yin Wong Ophthalmology