advertisement

Editors Selection IGR 9-4

Imaging: HRT reference planes

Comment by Linda Zangwill on:

21778 Analysis of HRT images: Comparison of reference planes, Poli A; Strouthidis NG; Ho TA et al., Investigative Ophthalmology and Visual Science, 2008; 49: 3970-3975

Find related abstracts

Identifying progressive optic disc change is one of the most important yet challenging aspects of glaucoma management. Not only do clinicians need to know whether glaucomatous progression has occurred, but how fast or slow the change is occurring to identify those with the highest risk of visual impairment due to glaucoma. Imaging instruments provide objective and reproducible measurements of optic disc and retinal nerve fiber layer, but there are few studies evaluating the longitudinal variability of these measurements. In this study, Poli et al. (1049) compare how different reference planes (RP) influence HRT rim area measurement variability and estimates of the rate of rim area change over time. The following five reference planes were evaluated, the standard reference plane (provided with HRT software), two reference planes defined at 320 microns posterior to the reference ring, a previously described experimental RP, and a new 'Moorfields RP'. This small series included 31 'converters, ocular hypertensives who developed reproducible visual field loss at follow-up, and 19 normal subjects who had undergone HRT imaging and field testing at least 'six times. The rate of change was estimated using linear regression techniques. A strength of this study is its inclusion of a group of normal subjects, but unfortunately the authors do not provide estimates of the rate of change, or variability for the normal group that can be directly compared to the converter group. Specificity rates for each RP are also not provided. In converters, the largest rate of change was found using the standard RP followed by the Moorfields RP, but differences among references planes were not statistically significant. In the converter group, the standard RP had significantly higher variability (the residuals standard deviation (RSD) of the rim area measurements, and slope/RSD) than the other RPs, and the authors suggest that the new Moorfields' RP offers an improvement over the standard reference plane. However, before one reference plane can be considered better than another, it is important to compare how many normal eyes remained stable (specificity), how many converters had significant change in rim area (sensitivity), and how many were progressing rapidly for each RP calculation. In addition, estimates of the variability (confidence limits and/or scatterplot) of the outcome measures (slope, slope/RSD, mean depth etc.) should be reported so that it can be determined whether the larger variability observed with the standard RP may be driven by a small number of eyes with very high variability, or a large number of eyes with moderate variability. Mixed effects models can be implemented to provide these estimates without the assumption of independence of the residuals over time. Although it is difficult to generalize the results of this small group of converters to the general clinic population, this study provides important information on how different reference planes may influence the estimates of the rate of rim area change using the HRT.

• ← Previous
• Next →