ASSESSMENT OF GLAUCOMATOUS VISUAL FIELD LOSS BY LOCAL FLUCTUATION OF VISUAL FIELD

Purpose:To compare the usefulness of local fluctuation and that of visual field parameters for assessing glaucomatous visual field loss. Subjects and methods:The subjects were 162 patients (323 eyes)who received perimetry two or more times at a mean age of 53 years. The mean observation period was 3 years and 10 months. Two types of criteria were established for these subjectsand eyes with normal visual fields were included in the analysis. We investigated whether an abnormal visual field could be predicted by specific parameters of the Cox proportional-hazards model using an abnormal visual field in each criterion as an objective variableand each measurement parame‑ ter obtained in a perimetry test as an explanatory variable. Measurement parameters includ‑ ed fixation loss false positive false negative mean deviation pattern standard deviation short-term fluctuation corrected pattern standard deviation glaucoma hemifield test local fluctuation (LF)and cluster. Results:With both types of criteriathe presence of LF at any timepoint had a significant correlation with the appearance of an abnormal visual field:crude hazard ratio of 4.3 (95% confidence interval (CI):3.2-5.9) and adjusted hazard ratio of 4.9 (95% CI:3.4-7.0)in crite‑ rion 1;and crude hazard ratio of 3.0 (95% CI:2.4-3.8) and adjusted hazard ratio of 3.5 (95% CI:2.6-4.8)in the other criterion according to the Cox proportional-hazards regression analysis. Conclusion:The presence of LF appears to be useful for predicting glaucomatous visual