COMPARISON OF METHODS TO DETECT VISUAL-FIELD PROGRESSION IN GLAUCOMA

Objective: The purpose of the study is to develop alternative statisti cal approaches for evaluating the trend of visual field series over ti me and to compare the results to human observers, Design: Retrospectiv e analysis of visual field results. Participants: Eighty-three eyes of 83 patients (phakic or pseudophakic) with open-angle glaucoma and 5 o r more eligible fields were included in the study, Intervention: Three experienced observers independently reviewed the field series to dete rmine stability or progression. Main Outcome Measures: The following a dditional methods to determine progression of visual field loss were u sed: (1) pointwise univariate regression analysis and a glaucoma chang e analysis; (2) univariate regression analysis on visual field indices mean deviation, corrected loss variance, and glaucoma pattern index; (3) pointwise multivariate regression analysis with fixed effects on p anel data; and (4) clusterwise multivariate regression analysis with f ixed effects on panel data, The results of different statistical metho ds were compared by determining the pairwise agreement (Cohen's weight ed kappa) between each technique and three experienced observers. Resu lts: Patients were observed for a mean (+/-standard deviation) of 5.6 (+/-1.4) years, The visual fields of 27 (33%) and 56 (67%) eyes were c onsidered to have progressed or remained stable, respectively, based o n agreement of at least 2 of 3 observers, Univariate regression analys is on visual field indices was not useful for detection of visual fiel d progression, Pointwise and clusterwise regression analyses with fixe d effects on panel data performed as well as pointwise univariate regr ession analysis compared with human observers (kappa = 0.52, 0.53, and 0.55, respectively). Both methods showed better agreement with human observers than with glaucoma change analysis (kappa = 0.41), Conclusio ns: A new statistical model, multivariate regression analyses with fix ed effects on panel data, is an appropriate method to evaluate the cou rse of visual field series over time and shows reasonable agreement wi th experienced observers and pointwise univariate regression analysis.