Misleading statistical calculations in far-advanced glaucomatous visual field loss

Area of Science:

• Ophthalmology
• Visual Science

Background:

• Glaucoma is a progressive optic neuropathy leading to characteristic visual field (VF) defects.
• Automated static perimetry is essential for diagnosing and monitoring glaucoma, relying on statistical analysis indices for interpretation.

Purpose of the Study:

• To assess the accuracy of statistical analysis indices in characterizing advanced glaucoma visual fields.
• To determine if summary and reliability indices can reliably detect abnormalities in end-stage glaucoma.

Main Methods:

• Retrospective observational case series of 16 eyes from 15 patients with end-stage glaucoma.
• Defined end-stage glaucoma by vertical cup-to-disc ratio (≥0.9), mean deviation (< -24 dB), and remaining VF island.
• Assessed collapse of statistical indices, including pattern deviation, corrected pattern standard deviation (CPSD), pattern standard deviation (PSD), short-term fluctuation (SF), glaucoma hemifield test (GHT), and patient reliability (false-negative rate).

Main Results:

• Nine of 16 VFs with misleading statistical calculations had normal pattern deviation probability plots.
• PSD, SF, and CPSD parameters were normal or near-normal in a significant proportion of eyes (4/16, 10/16, 5/16, respectively).
• Glaucoma hemifield test (GHT) results were not consistently abnormal, and low patient reliability (≥40% false-negative) was observed in 10 of 16 eyes.

Conclusions:

• Statistical indices are vital for interpreting automated static VFs.
• In end-stage glaucoma, both summary and reliability indices may fail to detect abnormalities, potentially misleading clinicians.
• Careful interpretation considering the limitations of statistical indices is crucial in advanced glaucomatous VF loss.