BACKGROUND: Clinical examinations of the contrast sensitivity (CS) are currently made by one-trial measurements of a restricted number of spatial frequencies around the peak of the CS function (CSF). These methods do not allow calculation of the confidence intervals of the single measurements, thus preventing examiners from evaluating the reliability of their data. We elaborated a new method of CS measurement at the peak of the CSF, aiming at economical clinical applicability together with enhanced reliability. Moreover, we tried to establish lower limits of normal CS that would allow classification of low test results as pathological. METHODS: A chart containing six columns of E optotypes of constant size and decreasing contrast (from about 16% to 0.5%) was used to perform six-fold contrast threshold determination (method of descending limits, forced choice). Optotype dimensions are intended to measure the CS at spatial frequencies corresponding to 3 and 6 cycles per degree (c/deg). The reference values were defined by testing 61 normal subjects: 31 "young" (16-35 years) and 30 "old" (56-81 years). RESULTS: The mean logCS found in the young group was 2.07 +/- 0.086 SD for 3 c/deg and 1.94 +/- 0.154 SD for 6 c/deg, while in the old group the mean logCS was 1.88 +/- 0.095 SD and 1.74 +/- 0.118 SD respectively. CONCLUSION: This simple method of measuring CS can provide a single CS value together with a confidence interval for the individual patient, thus allowing evaluation of measurement reliability.
BACKGROUND: Clinical examinations of the contrast sensitivity (CS) are currently made by one-trial measurements of a restricted number of spatial frequencies around the peak of the CS function (CSF). These methods do not allow calculation of the confidence intervals of the single measurements, thus preventing examiners from evaluating the reliability of their data. We elaborated a new method of CS measurement at the peak of the CSF, aiming at economical clinical applicability together with enhanced reliability. Moreover, we tried to establish lower limits of normal CS that would allow classification of low test results as pathological. METHODS: A chart containing six columns of E optotypes of constant size and decreasing contrast (from about 16% to 0.5%) was used to perform six-fold contrast threshold determination (method of descending limits, forced choice). Optotype dimensions are intended to measure the CS at spatial frequencies corresponding to 3 and 6 cycles per degree (c/deg). The reference values were defined by testing 61 normal subjects: 31 "young" (16-35 years) and 30 "old" (56-81 years). RESULTS: The mean logCS found in the young group was 2.07 +/- 0.086 SD for 3 c/deg and 1.94 +/- 0.154 SD for 6 c/deg, while in the old group the mean logCS was 1.88 +/- 0.095 SD and 1.74 +/- 0.118 SD respectively. CONCLUSION: This simple method of measuring CS can provide a single CS value together with a confidence interval for the individual patient, thus allowing evaluation of measurement reliability.