Performance of time-domain and spectral-domain Optical Coherence Tomography for glaucoma screening

• Published in: Acta ophthalmologica (Oxford, England) Vol. 90; no. 4; pp. 310 - 315
• Main Authors: Bengtsson, Boel, Andersson, Sabina, Heijl, Anders
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01.06.2012
• Subjects: Aged; Aged, 80 and over; Clinical Medicine; diagnosis; Exfoliation Syndrome - classification; Exfoliation Syndrome - diagnosis; False Positive Reactions; Female; glaucoma; Glaucoma, Open-Angle - classification; Glaucoma, Open-Angle - diagnosis; Humans; Intraocular Pressure; Klinisk medicin; Male; Medical and Health Sciences; Medicin och hälsovetenskap; Middle Aged; Nerve Fibers - pathology; Oftalmologi; Ophthalmology; Optic Disk - pathology; Optic Nerve Diseases - diagnosis; optical coherence tomography; Original; Predictive Value of Tests; Retinal Ganglion Cells - pathology; retinal nerve fibre layer thickness; Rural Population; screening; Sensitivity and Specificity; Sweden; Time Factors; Tomography, Optical Coherence - instrumentation; Sweden
• ISSN: 1755-375X; 1755-3768; 1755-3768
• DOI: 10.1111/j.1755-3768.2010.01977.x

. Purpose:  To investigate the measures of validity for selective or population screening of the time‐domain Stratus and the spectral‐domain Cirrus Optical Coherence Tomography (OCT) imaging techniques in a population‐based sample and in clinical glaucoma patients at different stages of glaucoma. Patients and methods:  A random sample of 307 subjects living in two rural areas in southern Sweden, and a random sample of 394 clinical glaucoma patients were selected. A large battery of examinations, including Stratus and Cirrus OCT was performed. OCT retinal nerve fibre layer (RNFL) thickness analyses for average thickness, quadrant and clock‐hour sectors were compared with normative significance limits available in the instruments. Results:  The population‐based sample included 129 healthy and nine glaucoma subjects, and the sample of clinical glaucoma patients included 138 patients. Specificity and positive predictive values were generally better with Stratus than for Cirrus, and sensitivity was generally better with Cirrus. With the average RNFL thickness parameter, Stratus reached 100% specificity and a positive predictive value of 100% and 68% sensitivity in the whole group of the clinical glaucoma patients, but sensitivity was only 28% among the earliest stage glaucoma patients. Sensitivity increased considerably when relying on the quadrant sector parameter, while specificity decreased only marginally. Conclusion:  Stratus, with high specificity and positive predictive values, seemed to be best of choice for screening purposes, while Cirrus, with high sensitivity, was the better choice for early detection.