advertisement
Lin et al. (1662) distributed six or more visual fields of forty patients who had participated in the AGIS study to eight glaucoma subspecialists and eight comprehensive ophthalmologists in three different formats: serial fields, STATPAC2 with GPA, and PROGRESSOR. They measured interobserver agreement within each group of observers and also compared the results of each to the Hodapp-Parrish-Anderson criteria. There was higher agreement among glaucoma subspecialists when using STATPAC2 GPA fields compared to other strategies and higher agreement with the Hodapp, Parrish, Anderson (HPA) criteria using STATPAC2 GPA. Interpretation time was also assessed and was shorter for PROGRESSOR when interpreted by glaucoma subspecialists and was shorter for STATPAC2 when interpreted by comprehensive ophthalmologists. The authors point out that the results may only be generalizable to patients with moderate to severe glaucoma since the beginning mean deviations averaged -10.4 dB and the mean final mean deviation average -13.4 dB. The study supports the use of computerized statistical analysis of visual fields for progression analysis.