advertisement

Editors Selection IGR 11-1

Structure and function measurements: HRT II versus HRT III

Comment by Marco Vizzeri on:

20332 Diagnostic ability of the heidelberg retina tomograph 3 for glaucoma, Ferreras A; Pablo LE; Pajarín AB et al., American Journal of Ophthalmology, 2008; 145: 354-359

Find related abstracts

Clinicians often face uncertainty in glaucoma management and rely on different technologies in an attempt to reduce this uncertainty. Since the armamentarium is continuously evolving, it is desirable before any new or updated device to be used in clinical practice that its diagnostic accuracy be first compared to that of previous devices. The HRT has recently been upgraded to the HRT-III version, providing a new diagnostic tool for data analysis, the GPS, and a new ethnic selectable database for MRA. Ferreras et al. (179) compared the diagnostic ability of HRT-III to that of the HRT-II version in 93 healthy and 90 glaucoma eyes, whose images were acquired using HRT2 but whose data was analyzed using both HRT-II and HRT-III. Eyes were classified as glaucomatous based on elevated IOP and abnormal SAP results. The parameters with the best area under the receiver operating characteristic curve (AUC) were the same for HRT-II and HRT-III, with the vertical cup-to-disc area ratio and the FSM and RB discriminant functions above all. Although the authors claimed that the diagnostic ability of HRT-III was somewhat superior to that of HRT-II for some of the stereometric parameters, the results showed rather similar AUCs and similar sensitivity and specificity values for MRA and GPS. In fact, the confidence intervals provided largely overlapped and, because no statistical analysis (such as the DeLong method) was applied to compare the AUCs obtained, the study did not convincingly demonstrate any HRT-III superiority. Also, it is known that study design and spectrum bias could significantly affect the performance of any diagnostic test.¹ In this study, by selecting healthy and diseased individuals based on SAP results, the diagnostic accuracy of both HRT-II and HRT-III was likely overestimated. Further studies are needed to compare HRT-II and HRT-III performance in glaucoma suspects, whose condition reflects the highest level of uncertainty for the examining clinician.

• ← Previous
• Next →