Glaucoma diagnostics

[caption id="attachment_8431" align="alignnone" width="370"] Mika Harju[/caption] Different kinds of optic nerve head (ONH) and retinal nerve fibre layer (RNFL) imaging techniques yield different results when evaluating patients for the presence of glaucoma, according to Mika Harju MD, Helsinki University Eye Hospital, Helsinki, Finland. In a study, he compared several imaging technologies as well as standard automated perimetry (SAP). RNFL photographs had the highest specificity and ONH photography had the highest sensitivity in patients referred for glaucoma testing. The study included 202 eyes of 101 patients. The basis for inclusion was at least one criterion suggestive of glaucoma – namely, ONH with cup/disc ratio 0.6 or more, a difference of more than 0.2 in the cup/disc ratio between their eyes, a mean intraocular pressure (IOP) higher than 21mmHg, and an ONH violation of the ‘ISNT-rule’. All eyes underwent examination with Humphrey Visual Field testing, ONH photography, RNFL photography, CIRRUS optical coherence tomography (OCT) ONH and RNFL imaging, and GdX scanning laser polarimetry. Each test results were assessed by three glaucoma specialists who classified them as glaucomatous, healthy or glaucoma suspects. The final determination of the presence or absence of glaucoma was made by five glaucoma experts using all data available from all examination methods. They found that, among the 202 eyes, 23 had glaucoma, 23 were glaucoma suspects, and 156 were healthy. Regarding each test in comparison to the consensus of the experts, GDx yielded 12 false positives and eight false negatives, and had a sensitivity of 62% and a specificity of 93%. OCT yielded eight false positives and 11 false negatives, and had a sensitivity of 52% and specificity of 99%. In addition, RNFL photography yielded only two false positives and nine false negatives, and had a sensitivity of 61% and a specificity of 99%. ONH photography yielded 13 false positives and four false negatives, and had a sensitivity of 83% and a specificity of 84%. Meanwhile, SAP yielded 15 false positives and 11 false negatives, and had a sensitivity of 52% and a specificity of 92%. Dr Harju noted that subjective evaluation by an ophthalmologist performed better than relying on cut-offs of software parameters only. That is, the sensitivity and specificity was only 67% and 82% respectively for Gdx, only 70% and 87% respectively for OCT, and only 57% and 86% 
for SAP. Mika Harju: 
mika.harju@hus.fi

Tags: diagnostics, glaucoma