Reproducibility of the Heidelberg retinal flowmeter in determining low perfusion areas in peripapillary retina

To evaluate the interobserver variability and retest reproducibility of confocal scanning laser Doppler flowmeter in measuring capillary perfusion of the peripapillary retina.

Blood flow measurements were performed in one eye of 10 normal subjects by two investigators on two different days (visits). Five separate measurements of the peripapillary blood flow parameters were recorded by each observer at each visit. The Heidelberg retina flowmeter was used to record capillary perfusion in a 2560x640 microm area of the superotemporal peripapillary region and pixel by pixel analysis was done from an area adjacent to the optic disc, with a minimum of 1600 pixels. The percentage of pixels with less than 1 arbitrary unit of flow (no flow) and 10, 25, 50, 75, 90th percentiles of flow values was calculated. Interobserver measurements were compared by paired t test. Intraclass correlations (ICC) were used to determine the interobserver variability and retest reproducibility of the measurements. Intrasession coefficients of variations (CV) were also calculated.

There were no statistically significant differences between the two observers for all measurements and between visits for the percentage of pixels with no flow. ICC was 66% (range 57.09%-77.1%) for pixels with no flow. For the 10, 25, 50, 75, 90th percentiles of flow the ICC was 63.07% (53.91%-77.81%), 71.3% (64.23%-80.85%), 72.61% (66.02%-78.96%), 65.86% (58.53%-74.77%), and 60.05% (54.34%-70.06%), respectively. CV was 16.59%, 11.47%, 9.32%, 9.03%, 11.58%, and 16.05% for the percentage of no flow pixels and the 10, 25, 50, 75, 90th percentiles of flow, respectively.

The Heidelberg retinal flowmeter allows reproducible measurements of all levels of capillary perfusion areas when pixel by pixel analysis is used.