Variables affecting test-retest variability of Heidelberg Retina Tomograph II stereometric parameters

To study the test-retest variability of stereometric parameters on the Heidelberg Retina Tomograph II, a new clinical instrument for glaucoma management.

In a cross-sectional study of 24 consecutive cases of glaucoma and 26 healthy subjects, Heidelberg Retina Tomograph II stereometric parameters from five consecutive images were obtained for one randomly selected eye of each subject. Test-retest variability was studied using three different statistical methods (coefficient of variance, intraclass correlation coefficient, and Cronbach alpha). The effect of age, diagnosis, linear cup/disc ratio, visual acuity, and refractive error on test-retest variability on HRT II was analyzed. The repeatability of Moorfields regression analysis and the baseline variability in the progression analysis software was also studied.

Using coefficient of variance, intraclass correlation coefficient, and Cronbach alpha, the test-retest variability was found to be lowest for mean cup depth, cup area, cup/disc area ratio, vertical cup/disc ratio and rim/disc area ratio, in that order. Test-retest variability had a significant correlation with age (r = 0.33, P = 0.019) and visual acuity (r = -0.46, P = 0.005). Compared with eyes with astigmatism less than 1 D (mean coefficient of variance = 6.4 +/- 4.9), the test-retest variability was higher (P = 0.044) in eyes with astigmatism more than 1D (mean coefficient of variance = 20.0 +/- 22.6). Moorfields regression analysis was inconsistent in 52% cases. The average baseline change in progression analysis software was 0.076 +/- 0.081.

The test-retest variability of Heidelberg Retina Tomograph II stereometric parameters is comparable to that reported for the Heidelberg Retina Tomograph. Eyes with uncorrected astigmatism more than 1 D and poor visual acuity may have a higher variability of Heidelberg Retina Tomograph II stereometric parameters.