Correlation of automated visual field parameters and peripapillary nerve fiber layer thickness as measured by scanning laser polarimetry

Purpose: To correlate Humphrey visual field mean sensitivity and peripapill ary nerve fiber layer thickness as measured by scanning laser polarimetry. Methods: The authors studied 54 eyes of 34 patients who visited a universit y-based glaucoma clinic and had undergone scanning laser polarimetry and Hu mphrey perimetry within 6 months. The study population included normal pati ents and those with glaucoma, ocular hypertension, and glaucoma suspect. Th e authors correlated visual field sensitivity with peripapillary nerve fibe r thickness, and visual field mean deviation with the average deviation fro m the normal nerve fiber layer thickness. They also correlated the visual f ield mean deviation with all available GDx Nerve Fiber Analyzer parameters. Results: The visual field mean sensitivity and deviation showed a bilinear correlation to peripapillary nerve fiber layer thickness. The visual field mean sensitivity changed little when the nerve fiber layer thickness was gr eater than 70 mu m. The nerve fiber layer thickness below this level was as sociated with a rapid decrease in the visual field sensitivity. Similarly, the visual field mean deviation was close to 0 dB when the nerve fiber laye r was within -10 mu m of the normal value; below this thickness, the mean d eviation became substantially more negative. There was a large individual v ariability around the bilinear fit. Of the scanning laser polarimetry param eters, a calculated index, referred to as the number, had the highest corre lation with the Humphrey mean deviation. Conclusion: The bilinear correlation and its variability between the scanni ng laser polarimetry and visual field parameters make it difficult to predi ct the result of one from the other. In general, the correlation between th e two is better when there is a significant visual field defect than when t he visual field is close to normal.