Comparative study of new imaging technologies for the diagnosis of glaucoma

DESIGN: A within patient multi-centre comparative study of the diagnostic performance of three automated imaging technologies for glaucoma diagnosis. Patients referred to secondary care eye services with suspected glaucoma will be approached. Each consented participant will undergo testing with the three technologies. Each test produces a diagnosis of glaucoma (yes/no) without observer input. The order of testing will be randomly selected for each participant. Then, the participants will be examined by an experienced glaucoma clinician who will perform a comprehensive examination including biomicroscopic slit lamp examination of the optic nerve and visual field testing (with Humphrey SITA 24-2 strategy) and provide the reference standard masked to the results of the imaging technologies. SETTING: NHS secondary care eye services. Four UK centres have been approached and agreed to participate. TARGET POPULATION: Adult patients referred from community optometrists to hospital eye services with suspected glaucoma, including those with ocular hypertension. HEALTH TECHNOLOGIES BEING ASSESSED: (i) HRT-III: Confocal laser scanning imaging technology, employed by the Heidelberg Retina Tomograph exploits the principle of confocal laser scanning to allow quantitative structural information of the optic disc anatomy. (ii) GDx-ECC: Scanning Laser Polarimetry measures the RNFL thickness. It is based on the birefringent properties of the RNFL. (iii) SD-OCT: Spectral Domain Optical Coherence Tomography is an optical imaging technique capable of providing high resolution, cross-sectional, imaging of the human retina and quantifies the thickness of the RNFL. MEASUREMENT OF COSTS AND OUTCOMES: For each instrument (HRT, GDx-ECC and OCT) the following outcomes will be measured, Primary diagnostic performance outcomes: sensitivity, specificity. Secondary diagnostic performance outcomes: ROC curves, diagnostic odds ratio, likelihood ratio, applicability, proportion of indeterminate tests. The overall diagnostic performance of combinations of these three technologies will also be evaluated, and their relative performance. Additionally, the diagnostic performance of the tests (and corresponding combinations) will also be assessed according to the spectrum of glaucoma (mild, moderate, severe), as defined by the specialist ophthalmologist and the impact of including a measure of intra-ocular pressure into the triage test will be investigated. Modelling (including economic modelling) will be used to determine which test or combination of tests would be effective and cost-effective compared with current practice. The results of the modelling will be diagnostic performance, cost and diagnostic outcomes and incremental cost per quality adjusted life year (QALY) gained.