Precision of the retinal tomograph as a screening device.

Purpose: To assess the diagnostic accuracy of the Heidelberg Retinal Tomograph 3 (HRT3) as a screening device in comparison with the reference standard of Octopus standard automated perimetry results (SAP) combined with clinical findings.

Methods: All patients underwent screening examinations and investigations within a single day. Abnormal screening results were classified as follows: The HRT3: Either "borderline" or "outside normal limits" using the global Moorfields classification (MFC); SAP and clinical exam: A mean defect>2.4 dB or "outside normal limits" clear text analysis of SAP; and one of the following i) IOP>21 mmHg, ii) Van Herrick<¼, iii) cup disc ratio>0.55, iv) optic nerve head abnormality, v) narrow iridocorneal angle or vi) evidence of peripheral anterior synechiae on gonioscopy.

Results: The mean age of the participants was 59.9 years (±14.8 [21, 91]). Twenty-three subjects (16%) were classified as abnormal on SAP and clinical exam. The HRT3 classification had a sensitivity of 30% (95% CI [16%, 51%]) with associated specificity of 58% (95% CI [49%, 66%]). Of the sixty subjects classified as borderline or outside normal limits with the HRT MFC global result, seven subjects were also abnormal according to SAP and clinical exam.

Conclusion: The results suggest that the HRT3 may not be suitable as a sole screening device; however, further investigation is necessary.