Diagnostic assessment of normal and pale optic nerve heads by confocal scanning laser ophthalmoscope and stereophotography

Purpose: To evaluate the diagnostic assessment of optic nerve heads (ONH) by stereophotography (SP) and Heidelberg retina tomograph (HRT) in patients with retinitis pigmentosa (RP). Methods: This study involved 53 consecutive patients with RP and 24 controls who underwent ONH examination by SP and HRT. On the basis of the appearance at SP, RP ONH were divided in normal-colored (24/53) and pale (29/53). The measurements of vertical ONH diameter and vertical cup/disc ratio (VCDR) by SP and HRT were compared between the groups by means of t test and Bland-Altman plots. Diagnosis, ONH pallor and size, HRT image quality (image standard deviation, SD) were also inspected as sources of differences in the results by means of a multivariate analysis. Results: Vertical diameter measurements were similar using SP and HRT, respectively (1.93 ± 0.50 and 1.80 ± 0.62 for normal-colored ONH, P=0.43; 2.02 ± 0.65 and 1.90 ± 0.40 for pale ONH, P=0.27; and 1.98 ± 0.55 and 1.84 ± 0.52 for controls, P=0.33). In normal-colored ONH, VCDR was 0.30 ± 0.11 at SP and 0.31 ± 0.21 at HRT (P=0.90); in controls, it was 0.29 ± 0.13 at SP and 0.30 ± 0.18 at HRT (P=0.77). In the group with pale ONH, VCDR was 0.34 ± 0.14 at SP and 0.53 ± 0.24 at HRT (P<0.001). HRT image quality was the only parameter which influenced the difference in VCDR estimates between SP and HRT (P=0.02). In patients with pale ONH, this difference was significantly higher, when images of acceptable quality were used for analysis (0.36 ± 0.15 vs. 0.15 ± 0.21 for images with good quality or better, P=0.03). Conclusion: In ONH diseases other than glaucoma, HRT may give misleading information when images of acceptable quality are considered for analysis. We therefore recommend the use of images with good quality or better (SD ≤30 μm) in these cases.