Inter-test reference height variability – a major error factor in Heidelberg Retina Tomography glaucoma progression analysis based on stereometric parameters changes

Abstract Purpose: To investigate the role of reference height inter-test variability upon the variability of the stereometric parameters. Materials and method: 204 glaucomatous patients underwent a complete ophthalmological exam, including Heidelberg Retina Tomography 3 (HRT-3). The exclusion criteria were optic disc or retinal pathology that might interfere with the detection of glaucoma progression, TSD >30μm. 4 sets of data were taken during the HRT-3 exam for each patient. Results: RH variability ranged between -198 and 187. Correlation analysis revealed a linear dependence between the inter-test variability of RH and stereometric parameters change. The most powerful correlations were observed for: RNFL Thickness (r=0.756, p<0.001), Rim Area (r=0.662, p<0.001), C/D Area Ratio (r=-0.663, p<0.001). The least correlated were Height Variation Contour (r=0.31) and Cup Shape Measure (r=0.07, p=0.3). When RH variability did not exceed 25μm, the correlations with stereometric parameters change were not statistically significant (for Rim Area, r=0.21, p>0.05, for C/D Area Ratio, r=-0.13, p=0.22, for RNFL Thickness r=0.06, p=0.52). Conclusions: For values >25μm, the variability of the RH is a major factor determining test/retest variability for RNFL Thickness, Rim Area, C/D Area, Rim Volume and Linear C/D. Inter-test variability of RH <25μm is an important criterion for the clinical relevance of stereometric parameters changes.