Data on over 14000 drivers from the 1967 California driver vision study have been reanalysed with a view to establishing their implications for driver vision standards. For the main analysis the sample was divided into four age groupings: under 25, 25-39, 40-54, and over 54. The most consistent result throughout the study has been the failure to find a direct relationship between poor visual performance and high accident rates for young and middle-aged drivers. For the over 54 age group, dynamic and static visual acuity showed the most consistent relationships with accident rates but for an individual driver their accident prediction value remained very low. A more detailed age analysis failed to define more precisely the age at which these relationships develop. No evidence was found to support the use of total visual field as a driver screening test. The results for two tests of night vision were regarded as inconclusive for the over 54 age group. For the same nominal standard of binocular static visual acuity, the ortho-rater screener was found to fail markedly fewer drivers than the Snellen wall chart. the implications of varying the cut-off scores for each test were investigated, and the suggestion is made that perceptual rather than sensory tests with greater accident predictive power would be needed before acceptable alternative screening methods could be specified for driver licensing purposes. (Author/publisher)
Abstract