PURPOSE: To compare the sensitivity and the specificity of the Heidelberg Retina Tomograph (HRT) classification of "Glaucoma" or "Normal" with that derived from clinical impression (CI) based on several parameters. DESIGN: Consecutive observational case series. METHODS: In a retrospective chart review of 200 left eyes of 200 consecutive patients referred to the Glaucoma Service Diagnostic Laboratory of the Wills Eye Hospital, we compared the HRT-based classification of "Glaucoma" or "Normal" with a CI classification of "Definite glaucoma," "Probably glaucoma," "Probably no glaucoma," and "No glaucoma." RESULTS: The HRT-based diagnosis of "Glaucoma" or "Normal" had an 86% sensitivity and 68% specificity when compared with a clinical impression (CI) of "Definite glaucoma," used as a strict gold standard definition of glaucoma; an 83% sensitivity and 57% specificity when the CI "Definite glaucoma" and "Probably glaucoma" were combined as a more liberal definition of glaucoma; a 76% sensitivity and 69% specificity when the CI "Definite glaucoma," "Probably glaucoma," and "Probably no glaucoma" were combined as the most liberal definition of glaucoma. The HRT diagnosis had an 86% sensitivity and 51% specificity when compared with the groups "Probably glaucoma," "Probably no glaucoma," and "No glaucoma" combined and considered as no glaucoma. CONCLUSIONS: In this study, HRT-based classification of "Glaucoma" or "Normal" was moderately sensitive, but not very specific when compared with the clinical impression as the gold standard. Clinicians should not rely on the HRT diagnosis alone, but should use it to supplement the impression based on an eye examination and other ancillary tests.
PURPOSE: To compare the sensitivity and the specificity of the Heidelberg Retina Tomograph (HRT) classification of "Glaucoma" or "Normal" with that derived from clinical impression (CI) based on several parameters. DESIGN: Consecutive observational case series. METHODS: In a retrospective chart review of 200 left eyes of 200 consecutive patients referred to the Glaucoma Service Diagnostic Laboratory of the Wills Eye Hospital, we compared the HRT-based classification of "Glaucoma" or "Normal" with a CI classification of "Definite glaucoma," "Probably glaucoma," "Probably no glaucoma," and "No glaucoma." RESULTS: The HRT-based diagnosis of "Glaucoma" or "Normal" had an 86% sensitivity and 68% specificity when compared with a clinical impression (CI) of "Definite glaucoma," used as a strict gold standard definition of glaucoma; an 83% sensitivity and 57% specificity when the CI "Definite glaucoma" and "Probably glaucoma" were combined as a more liberal definition of glaucoma; a 76% sensitivity and 69% specificity when the CI "Definite glaucoma," "Probably glaucoma," and "Probably no glaucoma" were combined as the most liberal definition of glaucoma. The HRT diagnosis had an 86% sensitivity and 51% specificity when compared with the groups "Probably glaucoma," "Probably no glaucoma," and "No glaucoma" combined and considered as no glaucoma. CONCLUSIONS: In this study, HRT-based classification of "Glaucoma" or "Normal" was moderately sensitive, but not very specific when compared with the clinical impression as the gold standard. Clinicians should not rely on the HRT diagnosis alone, but should use it to supplement the impression based on an eye examination and other ancillary tests.
Authors: Gadi Wollstein; Joel S Schuman; Lori L Price; Ali Aydin; Paul C Stark; Ellen Hertzmark; Edward Lai; Hiroshi Ishikawa; Cynthia Mattox; James G Fujimoto; Lelia A Paunescu Journal: Arch Ophthalmol Date: 2005-04