Analysis of visual field progression in glaucoma

Background - Despite the widespread use of computerised perimetry the diagnosis of visual field deterioration in following glaucoma patients over time remains particularly difficult. A new method of analysis using a novel graphical display of longitudinal field data is presented. Methods - A linear regression model of the luminance sensitivity at each stimulus location against time of follow up transforms the quantitative data from a series of fields into a colour coded form which illustrates the spatial configuration of change to aid the interpretation of field loss. The method of analysis and the developed computer software (PROGRESSOR) is described. Comparison with STAT-PAC-2 glaucoma change probability analysis is given including levels of agreement between the techniques using series of fields of 10 eyes from patients with normal tension glaucoma. Results - Examples of this new method compare well with STATPAC-2 analysis. The level of agreement between the techniques to separate progressing from stable retinal locations is good (kappa = 0.62; SE = 0.04). Conclusions - This new technique, which combines the change in perimetric sensitivity over time with colour coding of significant change into one image may provide an efficient method to detect true progression in glaucomatous field loss.