Reliability of the disk damage likelihood scale.

PURPOSE: To report the reliability of the glaucoma disk damage likelihood scale (DDLS) in comparison to the Armaly cup/disk ratio by determining the interobserver and intraobserver agreement for optic disk stereo photographs and the interobserver agreement for in vivo patient measurements of the optic disk.
DESIGN: Observational case series.
METHODS: Optic disk photographs: 48 stereo pairs of optic nerve photographs were selected from patients with a spectrum of glaucomatous visual field loss. Two masked observers graded the optic disk photographs three times according to the DDLS and Armaly cup/disk ratio. Interobserver and intraobserver agreements were calculated using the test-retest method. Patient measurements: three observers performed in vivo patient measurements on 34 eyes of glaucoma clinic patients and made a single determination of the DDLS stage and Armaly cup/disk ratio, based on the indirect biomicroscopic examination. Level of interobserver agreement was tabulated.
RESULTS: Optic disk photographs: interobserver and intraobserver agreement for the vertical DDLS measurement was greater than for two determinations (clinical impression and measured) of the vertical Armaly cup/disk ratio (interobserver: 85% vs 68% and 74%, respectively; intraobserver grader 1: 97% vs 89% and 80%, grader 2: 99% vs 95% and 89%, respectively). In vivo patient measurements: the interobserver agreement for the DDLS and Armaly cup/disk ratio was similar (70.1% vs 67.6%, respectively).
CONCLUSIONS: For the stereo optic disk photographs, the inter- and intra-observer agreement for the DDLS is greater than the Armaly cup/disk ratio. For the in vivo patient measurements, the level of agreement for the DDLS and the Armaly cup/disk ratio is similar.