Repeatability and comparison of clinical techniques for anterior chamber angle assessment.

Abstract

To assess the repeatability of gonioscopy, van Herick method and anterior segment Optical Coherence Tomography (AS-OCT) and determine the agreement between these techniques within a community optometry setting. Gonioscopy, van Herick method and AS-OCT imaging were performed by an optometrist on two occasions, 1month apart, on 80 subjects aged over 40years recruited from community optometry practices. Anterior segment images were captured with a spectral domain OCT (Topcon 3D OCT-2000; wavelength 840nm) set to the Anterior Segment (AS) mode. Eyes were graded as open or occludable for each method. AS-OCT images from both visits were graded by a second optometrist masked to the gonioscopy and van Herick method results, and the visit on which the images were acquired. Cohen's kappa (κ) was used to describe the intra-observer repeatability. Likelihood ratios, sensitivity and specificity of van Herick method and AS-OCT were calculated, using gonioscopy as the reference standard. Measurements were obtained from 80 eyes of 80 subjects. In four cases, AS-OCT images were un-gradable due to difficulty in locating the scleral spur. The repeatability of gonioscopy was fair κ=0.29, while that of the van Herick method (κ=0.54) and AS-OCT (κ=0.47) were somewhat better. The van Herick method showed good sensitivity (visit 1: 75%, visit 2: 69%) and high specificity (visit 1: 88%, visit 2: 96%). The sensitivity of AS-OCT was fair (visit 1: 46%, visit 2: 25%), but specificity was high (visit 1: 87%. visit 2: 89%). Intra-observer repeatability was better for van Herick method and for AS-OCT than for gonioscopy, despite the latter being considered the gold standard method. The van Herick method appeared to be more sensitive than AS-OCT when identifying eyes at risk of angle closure. A standalone anterior segment OCT with a longer wavelength laser could afford better visualisation of the angle, and might therefore be expected to enable the examiner to make more precise classifications. These instruments are not widely used by optometrists in clinical practice in the UK at present.