Objectives: To determine the potential of optic nerve head tomography [Heidelberg Retina Tomograph (HRT)] and scanning laser polarimetry (GDx) for identifying patients with glaucomatous visual field loss.
Design: Examinations were performed with the HRT, GDx and Humphrey Field Analyzer (HFA). Glaucoma was defined by the presence of a field defect. Patients within the cross-sectional groups underwent a single examination, whereas patients in the longitudinal groups were examined 6 monthly, for an average of 3.5 years.
Setting: Manchester Royal Eye Hospital, UK.
Participants: Patients with primary open angle glaucoma (POAG) or who were at risk of developing glaucoma.
Interventions: The diagnostic accuracies of the HRT and GDx were compared; specificity was set at 95%. The rate of change was determined by linear regression. To estimate the clinical application of the instruments, the proportion of an unselected group of patients on whom the examinations could be performed was calculated. Additionally, the time taken to perform and process each examination was measured.
Main outcome measures: The ability of the techniques to identify cases showing deterioration. The level of agreement and applicability of the techniques. Time taken to perform and process each examination.
Results: From the cross-sectional group, the maximum sensitivities of the HRT and GDx were 59% and 45%, respectively (at 95% specificity). From the two longitudinal cohorts, the level of agreement between the three instruments for identification of the development and deterioration of POAG was low. The applicability of the techniques was 80% (HRT), 88% (GDx) and 98% (HFA). The length of time to perform a full examination with each instrument was 12.3, 11.8 and 28.3 minutes, respectively. Agreement of HRT and GDx parameters between and within observers was largely good.
Conclusions: There is poor agreement for detection of glaucoma between the HFA, HRT and GDx. The techniques are amenable to use in the clinical environment, but no single examination has sufficient diagnostic precision to be used in isolation; also, the imaging techniques were not universally applicable. Neither the HRT nor GDx should be viewed as a replacement for visual field examination. Further research is needed into why most patients within the longitudinal arms of the study showed very little deterioration and into determining aspects of the structure versus function relationship in glaucoma that may explain why any one technique fails to detect a proportion of cases.