THE STATISTICAL INTERPRETATION OF BLUE-ON-YELLOW VISUAL-FIELD LOSS

Purpose. To evaluate short-wavelength-sensitive perimetry in the detec tion of glaucomatous field loss. Methods. The sample consisted of 27 n ormal subjects, 24 patients with primary open angle glaucoma (POAG), a nd 27 patients with ocular hypertension (OHT). Blue-on-yellow (B-Y) an d standard (W-W) perimetry was undertaken with a modified Humphrey Fie ld Analyzer 640 on one eye of each subject and patient. The B-Y data w ere corrected for individual ocular media absorption. Results were com pared to an age-matched normal database of 50 subjects (age range, 60 to 82 years; mean age, 70.0 years; SD, 6.4 years). Visual field indice s and total and pattern deviation probability maps were calculated for both W-W and B-Y fields. Results. The B-Y normal database exhibited i ncreased between-subject variability compared to the W-W normal databa se (P < 0.001). The greater variability increased with the increase in eccentricity (P < 0.001) and with the increase in age (P = 0.032). Al l patients with POAG exhibited B-Y field loss; 11 demonstrated greater B-Y loss than the corresponding W-W field. In advanced POAG, the B-Y and the W-W fields were similar. Twenty-five of the 27 normal subjects exhibited normal B-Y fields. Five of the 27 patients with OHT manifes ted B-Y focal abnormality and a normal W-W field: in two, W-W focal lo ss subsequently developed. Conclusions. Short-wavelength-sensitive per imetry can identify visual field loss before that detected by W-W peri metry. However, the increased between-subject variability necessitates stringent statistical analysis in the definition of abnormality.