Abstract
Purpose::
Several longitudinal studies have shown that new techniques like frequency doubling perimetry (FDT) and the nerve fiber analyzer (GDx) are able to predict glaucomatous visual field loss as assessed by standard automated perimetry (SAP). The results of these studies are often erroneously interpreted as if FDT/GDx are able to detect glaucoma earlier than SAP. Due to their design, however, these studies have a selection bias towards negative SAP findings: at baseline patients included in these studies may have normal or abnormal FDT/GDx test results whereas they all have normal baseline SAP findings. We investigated whether FDT and GDx are able to detect glaucoma earlier than SAP by following prospectively a cohort of glaucoma suspect patients with normal baseline test results for all tests.
Methods::
70 glaucoma suspect patients were followed prospectively for 4 years with SAP, FDT and GDx in a clinical setting. All patients had normal baseline test results for SAP, FDT and GDx. After the follow-up period, the numbers of patients who converted (changed from normal test result at baseline to reproducible abnormal test result during follow-up) on either technique were counted and compared. Cut-off point for FDT was >1 depressed test point P<0.01 in the total deviation probability plot; cut-off point for GDx was the Number >29. FDT was applied in full-threshold mode.
Results::
Of the 70 glaucoma suspect patients, 3 converted on FDT, 14 converted on GDx and 6 on SAP. These proportions are significantly different for GDx versus SAP (P=0.033) and GDx versus FDT (P=0.002) but not for FDT versus SAP (P=0.256).
Conclusions::
GDx is able to detect glaucoma earlier than SAP but FDT is not.
Keywords: perimetry • visual fields • imaging/image analysis: clinical