Contrast sensitivity and other vision tests in the optic neuritis treatment trial

To determine the intercorrelation, prevalence of abnormality, and incremental detection value of vision tests in optic neuritis.

We calculated the linear correlation of paired vision tests and prevalence of abnormal test values from baseline and six-month measurements of Snellen visual acuity, Pelli-Robson contrast sensitivity, Humphrey Field Analyzer mean deviation, and Farnsworth-Munsell 100-hue color vision in 438 patients entered in the Optic Neuritis Treatment Trial from 1988 to 1991. The incremental detection value of nonvisual acuity tests was defined as their frequency of abnormality when visual acuity was 20/20 or better.

All four vision-test results were highly intercorrelated at baseline and at six months. At baseline, contrast sensitivity had the highest prevalence of abnormality, but all vision tests were so often abnormal that differences were not clinically relevant. At six months, when visual recovery had occurred, contrast sensitivity was most often abnormal (2.2 X visual acuity; 1.8 X mean deviation; 1.5 X Farnsworth-Munsell 100-hue color vision test); when contrast sensitivity, mean deviation, or Farnsworth-Munsell 100-hue color vision was normal, visual acuity was 20/25 or better in 98% of patients.

The high intercorrelation of four vision tests suggests that optic neuritis affects a broad range of visual functions. Among non-visual acuity tests, Pelli-Robson contrast sensitivity proved to be a particularly practical and sensitive indicator of visual dysfunction in optic neuritis.