Comparison of the diagnostic capability of the Heidelberg Retina Tomographs 2 and 3 for glaucoma in the Indian population

Abstract

Purpose: To compare the diagnostic capability of the Heidelberg Retina Tomograph (HRT) 2 and 3 (Heidelberg Engineering, GmBH, Dossenheim, Germany) for glaucoma in an Indian population.

Design: Evaluation of a diagnostic technology.

Participants: Ninety-eight glaucoma subjects and 79 normal controls.

Methods: All participants underwent imaging with HRT2. Heidelberg Retina Tomograph 2 examinations were exported to HRT3 software. The stereometric parameters of HRT2 and HRT3 were compared. The diagnostic capability of Moorfields Regression Analysis (MRA) in the HRT2, HRT3 with and without ethnicity correction, and Glaucoma Probability Score (GPS) of HRT3 were compared. Analysis was done with the borderline results of MRA and GPS considered as normal to derive the most specific criteria and as abnormal to derive the least specific criteria.

Main outcome measures: Sensitivity, specificity, and area under the receiver operating characteristic curve (AUC).

Results: On standard automated perimetry (SAP), the mean deviation (mean+/-standard deviation) of the glaucoma and normal groups were -7.3+/-6.7 dB and -0.4+/-1.1 dB, respectively (P<0.001). The AUC for the individual stereometric parameters of HRT2 were similar to that of HRT3. The sensitivity and specificity (most specific criteria) were 56.1% and 96.2%, respectively, for MRA of HRT2 and 56.1% and 91.1%, respectively, for MRA of HRT3. The sensitivity and specificity (least specific criteria) were 88.8% and 88.6%, respectively, for MRA of HRT2 and 70.4% and 81%, respectively, for MRA of HRT3. By using the Indian specific normative database of HRT3, the sensitivity and specificity were 37.8% and 97.5%, respectively, for the most specific criteria, and 55.1% and 89.9%, respectively, for the least specific criteria. Sensitivity and specificity of GPS were 86.7% and 68.3%, respectively, for the most specific criteria, and 93.9% and 39.2%, respectively, for the least specific criteria.

Conclusions: In this cohort, the diagnostic capability of HRT2 MRA was similar to that of HRT3 MRA; ethnicity correction did not improve results; GPS was more sensitive but less specific than MRA in diagnosing glaucoma.