Clinical Evaluation of a Novel Population-Based Regression Analysis for Detecting Glaucomatous Visual Field Progression

Abstract

The population-based regression analysis seems to be superior to the trend analysis in detecting VF progression in glaucoma, and may eliminate the drawbacks of the event analysis. Further, it may assist the clinician in the evaluation of VF series and may allow better visualization of the correlation between function and structure owing to VF clusters.

Detection of Early Glaucomatous Progression With Octopus Cluster Trend Analysis

“…12,13,16 In a retrospective study performed on 240 consecutive visual field series with at least 9 consecutive tests, CCTA markedly increased sensitivity of detection of progression, compared with event analysis. 12 Trend analysis, however, requires several tests for optimal performance, which is not necessarily feasible in routine clinical practice when early visual field progression is investigated.…”

“…When the trend analysis results in a slope, which is not significantly different from that of the software's stable glaucoma reference population, the cluster is labeled as unaltered, but when it is significant, the cluster is labeled as belonging to one of the following categories: (1) significant worsening at P < 1% level; (2) significant worsening at P < 5% level; (3) significant improvement at P < 1% level; and (4) significant improvement at P < 5% level. When the long-term fluctuation is higher than that of the software's reference population 12,13 and the trend is not significant, the cluster is labeled as (5) significantly increased measurement variability at P < 1% or (6) at P < 5% level. When the software is not able to detect any further decrease of cluster MD because of 0 perimetric sensitivity in the cluster, the cluster is labeled as (7) floor effect.…”

“…4,11 This distribution made it possible to arrange functionally related test points into individual clusters. [11][12][13][14][15][16] The use of separate clusters allows separate analysis of functional progression for each cluster, based on trend analysis of the cluster MD value. 16 Cluster Trend Analysis (CTA) may be more sensitive to true changes than rate of global MD change or separate evaluation of the cluster's individual test point sensitivity values.…”

“…[17][18][19] It focuses on localized changes that may have minimal influence on global MD, but at the same time it minimizes the influence of background noise and the between-point dispersion. [11][12][13] In addition, as significant improvement and significantly increased long-term fluctuation compared with the data of the software's stable glaucoma reference population are also indicated, 11,16 correct clinical decision making on cluster progression or variability is strongly supported by the software.…”

Evaluation of Octopus Polar Trend Analysis for Detection of Glaucomatous Progression

“…Unlike corrected cluster trend analysis, which employs trend analysis to investigate the progression of local defect values averaged for functionally connected test points (clusters) of the visual field (6), PTA is based on pointwise regression analysis (6, 13-15). Cluster analysis minimizes the influence of background noise and the between-point dispersion (6, 13–15, 21). These advantages, however, are not available for pointwise regression analysis.…”