The retrospective study compared guided progression analysis (GPA), pointwise linear regression (PLR), and the glaucoma rate index (GRI) for estimating visual field progression and suggests GRI was able to detect long-term progression in primary open-angle glaucoma earlier than PLR or GPA.
Using data collected from May 1988 to November 2004, investigators obtained information on 726 eye of 567 consecutive patients with open-angle glaucoma for the current analysis. Of the aforementioned cohort, 257 eyes of 176 patients treated at the Glaucoma Division of Stein Eye Institute at UCLA and 472 eyes of 391 patients were obtained from the Advanced Glaucoma Intervention Study.
Of note, all patients included in the study had 6 or more longitudinal visual field measurements and a minimum follow-up of 3 years. Investigators pointed out visual field measurements were considered reliable if they had less than 15% false-positive rates and less than 30% fixation losses and false-negative rates according to previous criteria.
Outcomes of interest for the current study included the proportion of Proportion of VF series detected as progressing, estimates of false-positive proportions, time to detect progression, and agreement among measures. Investigators also noted a subgroup analysis was performed in patients with likely visual field progression and likely visual field stability.
In regard to the 567 patients included, 52% were white, 300 were female, and the mean age of the cohort was 65.9 (9.7) years. Additionally, median follow-up for the study cohort was 8.9 (IQR, 7.3-10.4) years and the median number of visual fields obtained was 14 (IQR, 12-17)—in total, 4410 visual field measurements were part of the analysis.
When comparing the 3 pointwise methods, the proportion of eyes labeled as progressing was 27.7% using GPA, 33.5% using PLR, and 52.9% using GRI. Results indicated a pairwise difference of 20% (95% CI, 17%-23%) for GRI versus PLR, 25% for GRI versus GPA and 6% for PLR versus GPA (P <.001 for all).
Median time to progression was 8.8 (IQR, 2.4-10.5) years with GRI compared to greater than 16 years for both GPA and PLR. Additionally, investigators noted the results of Cox proportional hazards regression indicated hazard ratio of visual field progression for GRI versus PLR was 11.3 (95% CI, 9.2-13.7), 18.1 (95% CI, 14.5-22.6) for GRI versus GPA, and 1.5 (95% CI, 1.3-1.9) for PLR versus GPA (P < .001 for all).
Furthermore, in 156 patients labeled as likely progression, the proportion of progressing eyes were 73.7% for GPA, 81.4% for PLR, and 92.9% for GRI. Results indicate a pairwise difference of 11.5% (95% CI, 7.4%-17.6%; P <.001) for GRI versus GPA, 19.2% (95% CI, 12.6%-26.4%; P <.001) for GRI versus GPA, and 7.7% (95% CI, 0.3%-15.7%; P=.08) for PLR versus GPA.
Investigators noted multiple limitations within their study, including the retrospective and observational design. Other limitations included irregular intervals of visual field testing in the UCLA group, including patient with a wide range of disease severity and long-term follow-up could improve clinical relevance, and low sensitivity AGIS score restricted identification of patients with true-positive findings.
This study, titled “Pointwise Methods to Measure Long-term Visual Field Progression in Glaucoma,” was published in JAMA Ophthalmology.