Longitudinal Analysis of Reliability Indices in Glaucoma Patients and Normal Controls

Abstract

 

Abstract:

 

 

To study serial reliability indices of static automated perimetry in a longitudinal prospective study of open–angle glaucoma patients and normal controls.

 

 

The sample contained 197 patients and 113 controls (mean age: 63 ± 12 and 50 ± 12 years, respectively) followed in our laboratory over 7.5 ± 2.9 and 8.0 ± 4.1 years, respectively. All subjects were tested with the full threshold 30–2 programme of the Humphrey Field Analyser at regular intervals of 4–6 months. While patients had both eyes tested, the study eye was always tested first. Controls were tested only in the study eye. In a subset of patients, a subjective reliability score (from 1 (excellent) to poor (5)) based primarily on attentiveness was given to each examination independent of the perimetric reliability indices.

 

 

A total of 8260 visual fields were analysed, with a mean of 19 and 13 examinations respectively in the study eyes of patients and controls. The respective overall mean false positive (FPR), false negative (FNR) and fixation loss (FLR) rates were 3.3%, 5.0% and 6.7% in patients and 4.1%, 1.6% and 7.4% in controls. There were no systematic changes in any of the indices over time (Figure). The subjective score correlated highly with FLR, but not FPR or FNR. In both groups there was a weak but consistent association between FLR and FPR, while in patients there was a strong inverse relationship between Mean Deviation and FNR. The mean between–eye difference in all indices was < 1%.

 

 

FPR, FNR and FLR in this study were substantially lower than in previous reports. There was little evidence of longitudinal change. The magnitude and longitudinal pattern of FPR and FLR were similar in patients and controls, however FNR was always higher in patients.

 

 

 

Keywords: visual fields • perimetry