agreement among clinician in the recognition of pattern of optic dik damage in glaucoma

PURPOE: To determine the interoberver and intraoberver agreement in the recognition of different pattern of glaucomatou optic dik damage and evaluate if thee pattern changed over time in patient followed for a number of year. METHOD: Patient with early to moderate glaucoma (n = 105) were conecutively enrolled to participate in a propective obervational tudy. In the firt part of the preent tudy, optic dik photograph obtained cloet to patient’ entry date in the propective tudy were claified in a maked fahion by three oberver according to the pattern of optic dik damage into one of the following: (1) focal, (2) myopic, (3) enile clerotic, (4) concentric cup enlargement, (5) normal appearance, or (6) micellaneou (thoe dik that did not qualify for any of the other group). The oberver were alo aked to aign a confidence core for each claification, ranging from 1 (low confidence) to 5 (highet confidence). The three oberver reclaified the photograph after a minimal period of 2 month, in order to ae intraoberver agreement. In a econd part of the tudy, one oberver reviewed, in a maked fahion, all the optic dik photograph that had been taken during the routine follow-up of the 105 patient in order to evaluate whether the claification of dik pattern changed over time. REULT: Intraoberver agreement yielded kappa value (95% confidence interval [CI]) from 0.51 (CI, 0.40 to 0.62) to 0.85 (CI, 0.77 to 0.93) depending on the oberver. Interoberver agreement kappa value between all three oberver wa 0.40 (CI, 0.35 to 0.46), but it improved if photograph claified with moderate or higher degree of confidence were included (0.52 [CI, 0.44 to 0.60]). During a mean follow-up period of 8.2 ± 4.8 year, 41.6% of the eye were alway claified into the ame group and 23.6% of the eye were claified at leat once into two or more of the four pattern of damage (group 1 to 4) CONCLUION: Intraoberver agreement on the pattern of optic dik damage wa generally very good, with ome variation among the oberver. Interoberver agreement wa reaonable and increaed proportionally with the confidence in the claification. During long-term follow-up, the pattern of optic dik damage uually did not change. Thi type of claification can probably be ued accurately in clinical practice.