Mild Learning Effect of Short-wavelength Automated Perimetry Using SITA Program

Abstract

To evaluate the learning effect at Short-wavelength Automated Perimetry (SWAP) using the Swedish Interactive Threshold Algorithm (SITA) program over the central 24 degrees on patients with ocular hypertension experienced with standard automated perimetry. Twenty-seven patients underwent 5 SITA SWAP tests at intervals of 5+/-2 days in a randomized eye. Learning effect was defined as an improvement in results for duration, perimetric indices, foveal sensitivity, Glaucoma Hemifield Test, number of points with a P<5% and <1% in the total, and pattern deviation maps, total, central, paracentral, peripheral, and quadrant sensitivity. Learning effect was shown for foveal sensitivity (P=0.006, analysis of variance), which was significantly different only between the first (23.0 dB+/-4.1 dB) and the last test (24.7 dB+/-3.8 dB, P=0.003, t test) and was not affected by any demographic or ophthalmic characteristics of the population. All the other parameters did not show any significant learning effect. SITA SWAP is affected by a small learning effect interfering only with the first test, as opposite to full-threshold SWAP program, which previously showed severe learning artefacts. This fact may induce reconsideration of the clinical utilization of SWAP for the early diagnosis of glaucoma.