Abstracts Category : Other

Variability in perimetry and optical coherence tomography: its importance in glaucoma detection, diagnosis and management

by John Graham Pearce

The detection, diagnosis and monitoring of glaucoma requires numerous investigative techniques and strategies in order to provide definitive diagnosis and management of glaucoma. Two investigative techniques used extensively for this purpose today are standard automated perimetry (SAP) and spectral domain optical coherence tomography (SD-OCT). The presence of test-retest variability (TRV) for repeat measures in these instruments can impact on our ability to differentiate TRV from true progressive change. In this thesis, we examine two popular devices which, at present, have little or no published TRV criteria available. The instruments under consideration are the Medmont M700 automated perimeter and the Topcon 3D OCT-2000 SD-OCT. The purpose of this study is to quantify TRV for both instruments in both healthy and glaucoma patients, and identify any modifiable external components of TRV. Chapters 1 and 2 determine the test-retest variability (TRV) of both instruments in glaucoma patients. We were able to develop event based progression criteria which could be easily applied by clinicians using these instruments. We also established that TRV for several summary measures, structural and functional, was dependent on the level of glaucomatous damage. We reported that for the Medmont, retest variance did not increase with eccentricity when comparing points of equal decibel value in the central 10 of the 30-2 test with points of equal decibel value in the outer 20 of the same test. This finding had not been reported previously, and was counter to that of other studies (involving healthy eyes) which had reported that variability increased with eccentricity. There are few published studies which have examined diurnal variations in SAP and SD-OCT summary measures, and none that had investigated diurnal variations in TRV in either instrument. In Chapter 3, we therefore investigated whether diurnal variation existed in either summary measures or TRV. Whilst we found no detectable changes to mean summary measures over the course of a normal working day, we were able to detect statistically significant diurnal variations in TRV for several SAP and SD-OCT summary measures. This was now evidence to suggest that reduced TRV, and therefore enhanced progression detection, might be available by testing subjects on both instruments at consistent and specified times of day. In Appendices 1 and 2, we investigated whether any circadian variations in SAP and SD-OCT summary measures were detectable over a 24 hour period. We were unable to discern any significant differences in SAP or SD-OCT summary measures, nor were we able to discern any consistent cycles in any summary measures. We were able to detect statistically significant circadian variation in the peripapillary retinal nerve fibre layer using SD-OCT,