Abstract
Purpose: :
No standardized criteria to define progression using 10-2 visual field (VF) tests in glaucoma management currently exist. We investigated the performance of potential progression criteria for 10-2 VF tests using trend analysis.
Methods: :
We reviewed the 10-2 VF results of patients with severe glaucoma (24-2 VF MD < –12 dB) who performed ≥5 tests during follow-up. Global (MD and PSD) and localized (pointwise) rates of VF change (dB/yr) were calculated using linear regression providing slopes and P-values for each measured outcome based on 10-2 results. Progression (sensitivity) was defined using different combinations of steepness and P-values for the slopes, whereas false-positive results (1–specificity) were defined using the same criteria as those for progression but using positive slopes (i.e., VF "improvement"). For example, for a progression criterion with a slope <–1.0 dB/yr at P <0.05, a false-positive result was defined as >+1.0 dB/yr at P <0.05. The criterion with a combination of slopes and P-values leading to the highest likelihood ratio (LR= sensitivity/1–specificity) was deemed the one with best performance.
Results: :
We included 178 eyes of 146 patients with a mean of 10.6±7.7 10-2 VF tests spanning 8.3±4.6 years. The mean baseline 10-2 MD and PSD values were –12.1±7.8 and 8.5±4.2 dB, respectively. The average rates of 10-2 MD and PSD change over time were –0.52±0.7 and +0.13±0.4 dB/yr, respectively. When defining progression based on the presence of at least one VF location progressing <–1.0 dB/yr at P <0.01, 130 (73%) eyes were deemed as progressing with 40 (22%) false-positive results (LR=3.3). Using the same definition but requiring the presence of at least 2 VF locations in the same arcuate region of a hemifield ("comma-shaped") defect, 99 (55%) eyes progressed with 19 (10%) false-positives (LR=5.5). All other tested criteria lead to smaller LR values than the ones above.
Conclusions: :
In eyes with severe glaucoma, clinically significant VF deterioration in the central 10 degrees can be accurately detected using trend analysis. This is the first study to investigate progression criteria for 10-2 VFs using rates of change and to test their performance. These findings may be useful for monitoring glaucoma patients with severe functional loss.