Purchase this article with an account.
A. Sinapis, C. Bergin, D. Sinapis, R. A. Russell, R. Moosavi, C. Balian, G. M. Verdon-Roe, J. Flanagan, D. P. Crabb, D. F. Garway-Heath; Perimetry Instrument Comparison Study: Between Test Variability of Four Threshold Perimetry Tests in Healthy and ‘Glaucomatous’ Eyes (Interim Analysis). Invest. Ophthalmol. Vis. Sci. 2010;51(13):5506.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To consider interim results comparing the between test variability characteristics of the threshold strategies of four perimetry devices.
One-hundred and two, from a planned 320, participants have been tested in a four centre study of perimetry instruments (mean age=60 years, range 20 to 83 years). Selection criteria for patients were: (i) Heidelberg Retina Tomograph (HRT) Moorfields Regression Analysis (MRA) ‘Outside Normal Limits’; rim area >0.5mm2 and (ii) intraocular pressure (IOP) >21mmHg at referral; <26mmHg at time of testing. Healthy subjects had HRT MRA ‘Within Normal Limits’ and IOP ≤21mmHg. Thus, the reference standard for ‘glaucoma’ was independent of perimetry. One eye of each subject was tested on two occasions within 15 weeks with Standard Automated Perimetry (SAP, 24-2 SITA standard), Frequency Doubling Technology Perimeter (FDT, 24-2 ZEST), Heidelberg Edge Perimeter (HEP) and the Moorfields Motion Displacement Test (MDT). Test order was randomized. To evaluate the between-test variability of each instrument, Bland-Altman plots and repeatability coefficients (RCs) were obtained from pairs of measurements (Z-scored by the standard deviation of normative values) for the 32 between-device matched test locations.
The Bland-Altman plots suggested that between-test variability was lowest in HEP; while SAP, FDT and MDT appeared to have higher variability. The results were confirmed by the RC of each instrument. RCs for SAP, FDT, HEP and MDT were 0.37, 0.58, 0.21 and 0.45, respectively.
Repeated visual field measurements were found to be least variable, relative to between-individual normal variation, for HEP, followed by SAP, then MDT, then FDT. However, imprecision in estimating the normal ranges for some devices may explain the larger variability in these devices.Acknowledgement of Collaborators: L. Tanga, M. Michelessi, F. Oddone (Bietti Foundation, Rome, Italy)G. Sharpe, P. Artes (Dalhousie University, Halifax NS, Canada)
This PDF is available to Subscribers Only