Comparison of Size Modulation Standard Automated Perimetry and Conventional Standard Automated Perimetry with a 10-2 Test Program in Glaucoma Patients.

Purpose: This prospective observational study compared the performance of size modulation standard automated perimetry with the Octopus 600 10-2 test program, with stimulus size modulation during testing, based on stimulus intensity and conventional standard automated perimetry, with that of the Humphrey 10-2 test program in glaucoma patients.

Methods: Eighty-seven eyes of 87 glaucoma patients underwent size modulation standard automated perimetry with Dynamic strategy and conventional standard automated perimetry using the SITA standard strategy. The main outcome measures were global indices, point-wise threshold, visual defect size and depth, reliability indices, and test duration; these were compared between size modulation standard automated perimetry and conventional standard automated perimetry.

Results: Global indices and point-wise threshold values between size modulation standard automated perimetry and conventional standard automated perimetry were moderately to strongly correlated (p < 0.01). However, the correlation coefficient of point-wise threshold value for the central zone was significantly lower than that for the peripheral zone (χ2 > 33.40, p < 0.01). Better mean defect and point-wise threshold values were obtained with size modulation standard automated perimetry than with conventional standard automated perimetry, but the visual-field defect size was smaller (p < 0.01) and depth shallower (p < 0.01) on size modulation-standard automated perimetry than on conventional standard automated perimetry. The reliability indices, particularly the false-negative response, of size modulation standard automated perimetry were worse than those of conventional standard automated perimetry (p < 0.01). The test duration was 6.5% shorter with size modulation standard automated perimetry than with conventional standard automated perimetry (p = 0.02).

Conclusions: Global indices and the point-wise threshold value of the two testing modalities correlated well. However, the potential of a large stimulus presented at an area with a decreased sensitivity with size modulation standard automated perimetry could underestimate the actual threshold in the 10-2 test protocol, as compared with conventional standard automated perimetry.