Comparison of Swedish Interactive Threshold Algorithm and Full Threshold Algorithm for Glaucomatous Visual Field Loss

Abstract

To compare the prevalence of visual field loss, the sensitivity distribution, and the size and depth of glaucomatous visual field defects using the standard full threshold (FT) and the Swedish interactive threshold algorithm (SITA) standard (SS) procedures in patients with early or suspected glaucoma. Automated perimetry findings were retrospectively evaluated in 53 patients (105 eyes) with early or suspected glaucoma. The number of eyes judged to have glaucomatous visual field loss by SS (48 eyes) was significantly larger than what was found with FT (35 eyes), and 70 eyes were classified as pre-perimetric glaucoma. In these 70 eyes, there were many locations where the sensitivity was significantly higher with SS than with FT (intrasubject difference), and SS had less intersubject variability than FT at most locations. The cumulative decibel scores at the region of glaucomatous defects were larger with SS (206.2+/-103.3 dB) than with FT (162.1+/-87.5 dB) (p=0.02), which indicated that the depth of defects measured by SS was shallower than that by FT. The sizes of defects were significantly larger with SS (11.2+/-5.6) than with FT (9.7+/-5.1) (p<0.05). Glaucomatous defects were measured as being significantly shallower and larger with SS than with FT. In addition, the prevalence of visual field defect was higher with SS according to some of the criteria for glaucomatous visual field defects. These results might be related to the fact that SS strategy has a lower variability and to the Bayesian statistical properties of the SITA algorithm.