Abstract
Purpose :
Glaucoma patients often complain about poor vision in extreme (low or high, or suddenly changing) luminance conditions whereas visual performance with optimal luminance tends to be rated as normal. We developed a questionnaire with 41 luminance-specific questions to assess (1) if visual complaints of glaucoma patients are, compared to controls, indeed more pronounced in extreme luminance conditions and (2) if a combination of luminance-specific questions could differentiate glaucoma patients from controls.
Methods :
We sent a questionnaire to 221 glaucoma patients randomly selected from our outpatient perimetry database; controls were primarily their spouses. Questions were addressing visual performance under five conditions: (1) presumed optimal luminance (outdoor on a cloudy day), (2) low luminance (outdoor at night), (3) high luminance (outdoor on a sunny day), (4) sudden decrease in luminance, and (5) sudden increase. For each category, we selected the question that differentiated best between both groups. A decision tree (CHAID) was used to determine sensitivity and specificity of detecting glaucoma with the five selected questions.
Results :
We retrieved 178 questionnaires from the glaucoma patients (response rate 81%); 182 controls completed the questionnaire. Mean (SD) age of the patients and controls was 72.2 (10.0) and 65.7 (10.8) years, respectively (P<0.001). The median (IQR) Humphrey field analyzer mean deviation of the patients’ best eye was -4.7 (-2.0 to -11.3) dB. Glaucoma patients reported more complaints compared to controls in dark (50 versus 7%), light (35 versus 4%), light-to-dark (37 versus 4%) , and dark-to-light (34 versus 6%) conditions. Optimal luminance resulted in a smaller difference between the groups (13% versus 0%). The decision tree resulted in a sensitivity of 62% and specificity of 98%. Using an age-matched subgroup of controls, the specificity was 96%. The dark condition discriminated best, but all were contributing.
Conclusions :
Differences in visual complaints between glaucoma patients and controls are more pronounced in extreme luminance conditions than with optimal luminance. With only five luminance-specific questions, it is possible to differentiate glaucoma patients from controls with a good sensitivity and a high specificity. If confirmed in an independent group, these findings offer a novel approach to glaucoma screening.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.