Impact of better and worse eye damage on quality of life in advanced glaucoma

Abstract

The purpose of the study was to investigate the influence of VF and the VA on vision related quality of life (VRQoL) in advanced glaucoma. Subjects consist of 50 glaucoma patients with mean deviation (MD) less than -20 dB in at least one eye. Patients' VRQoL was assessed using the 'Sumi questionnaire'. The impact of seven visual measures on VRQoL were compared using principal component regression: MDs of better and worse eyes with 10-2 and 24-2 Humphrey VFs, LogMAR VAs of better and worse eyes and the Esterman score. The root mean of the squared prediction error (RMSE) was calculated using leave-one-out cross validation. Better eye summary measurements were much more influential on VRQoL than corresponding worse eye measurements and Esterman score in every VRQoL task. In conclusion, in advanced glaucoma, VF parameters of the better eye are important for the VRQoL of the patient.

Figure 1. 24-2 VF and 10-2 VF of sample cases.

Figure 1(a), Grayscales of case a (80 year old, male). MDs of the right eye were −29.6 (24-2 VF) dB and −26.5(10-2 VF) dB and those of the left eye were −4.0 (24-2 VF) and −2.0 dB (10-2 VF). Figure 1(b), Grayscales of case b (72 year old, male). MDs of the right eye were −10.0 (24-2 VF) dB and −0.7(10-2 VF) dB and those of the left eye were −27.7 (24-2 VF) and −21.5 dB (10-2 VF). Figure 1(c), Grayscales of case c, (78 year old, female). MDs of the right eye were −27.2 (24-2 VF) dB and −27.8(10-2 VF) dB and those of the left eye were −27.6 (24-2 VF) and −23.9 dB (10-2 VF). MD: mean deviation, VF: visual field.

Figure 2. Histogram of better-eye MD 24-2.

Figure 3. RMSE with first to seventh PCA components.

RMSE was obtained using the leave-one-out cross validation method and standard linear regression. Adding additional PCA components to the first component did not significantly decrease RMSE (Wilcoxon test, p ≥ 0.05). RMSE: root mean of the squared prediction error, PCA: principal component analysis.