Abstract
Purpose: :
Much research in glaucoma concerns the enigma of whether an eye is progressing or stable (Spry & Johnson 2002, Surv Ophthalmol 47:158–173). Predicting and preventing progression in eyes with early or suspected glaucoma is a key aim. Such studies require longitudinal data from a population of progressing eyes, but visual field progression for very early glaucoma in clinical studies is surprisingly rare.
Methods: :
The Perimetry and Psychophysics In Glaucoma (PPIG) project provides an unrivalled set of very high quality longitudinal data. To date, at least 7 years of annual test results from subjects with early glaucoma or glaucomatous optic neuropathy are available for 168 eyes from 84 subjects using standard automated perimetry (SAP), and 164 eyes from 82 of those subjects using short wavelength automated perimetry (SWAP). All subjects had some previous experience with perimetry before entering the study.
Results: :
For SAP, the mean sensitivity (MS) of the 52 non–blind spot locations, averaged over all 168 eyes, improved from 27.72dB in year 1 to a maximum of 28.24dB in year 5, before subsequently falling. The average number of progressing test locations (defined using pointwise linear regression as being a location progressing at –1dB/year over a five year period, significant at the 1% level) rose from 0.08 per eye over years 1–5, to 0.30 per eye over years 3–7. For SWAP, the average MS improved from 18.49dB in year 1 to a maximum of 20.02dB in year 6, before falling; and the average number of progressing locations rose from 0.21 per eye over years 1–5, to 0.50 per eye over years 3–7.
Conclusions: :
Practice (learning) effects and strict entry criteria cause the prevalence of progression in any clinical trial of early and suspected glaucoma subjects to be low until several years have passed. When only one test is conducted per year, practice effects occur for at least 6 years in the average subject. Because of aging, possible true progression, and subjects’ prior experience with perimetry, these results are actually likely to underestimate the true time period. The extended time period of this practice effect and the level of prior experience of the subjects must be taken into account, especially when conducting clinical trials.
Keywords: perimetry • visual fields • clinical (human) or epidemiologic studies: prevalence/incidence