Variability in repeated contrast sensitivity measures: implications for the individual patient

We measured contrast sensitivity at three spatial frequencies (3, 6, 12 cycles/degree) in nine observers repeatedly over four weeks to establish the level of practice effects and the variability to be expected within subjects on repeated testing. We found that the variability for the group as a whole was approximately three times that for any individual within the group. If the group norms were used to determine whether a patient's responses were abnormal, subjects with highest contrast sensitivity could he as many as 28 standard deviations from their own mean before they would be considered for referral. These and similar data on other tests are important in establishing statistical limits for measures of visual function, so that abnormalities can be detected and treatment of disease can be started as early as possible.