Purchase this article with an account.
Richard A. Russell, Rizwan Malik, David P. Crabb, Ananth C. Viswanathan, David F. Garway-Heath; Estimating the Relationship between Variability and Measured Sensitivity in VFs using Large-Scale Patient Data Series. Invest. Ophthalmol. Vis. Sci. 2011;52(14):5490.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Evaluation of visual field (VF) damage is largely based on pointwise sensitivity data from standard automated perimetry; however, measurements are highly variable (as shown by frequency-of-seeing and test-retest studies), which significantly impedes assessment of progression. The aim of this study was to characterize VF variability by level of sensitivity using a statistical method to quantify heteroscedasticity.
68,099 Humphrey 24-2 SITA Standard VFs from 8,252 patients (14,137 eyes) visiting Moorfields Eye Hospital from 1997-2009 were studied. Each eye’s VF series was analyzed using pointwise linear regression of sensitivity over time, with residuals (difference from fitted value) from each regression pooled according to the fitted sensitivities (to the nearest dB). Average variability was calculated as the median residual for each whole dB sensitivity level.
The median (interquartile range, IQR) patient age, follow-up and series length was 66 (55-74) years, 5.3 (3.1-7.5) years and 4 (3-8) VFs, respectively. The inferred variability is shown (cf. Artes et al. and Henson et al. estimates) in Figure 1. Overall, there is good agreement with Artes et al.; however, in the 35-40dB range, Artes et al. estimates are much smaller. This likely reflects the real-world nature of a clinical dataset with a higher rate of false-positive responses.
This study highlights a new approach for characterizing VF variability with sensitivity. A great strength of the method is that inference is based on thousands of clinic patients rather than the tens of subjects in test-retest studies. The results are important for helping clinicians to distinguish real VF progression from noise and will be used in models for glaucoma progression detection.
This PDF is available to Subscribers Only