Content area
Full Text
Neuropsychological assessment of patients with acquired brain injury is critical because such cases are common causes of enduring significant disability. Although neuropsychological tests are designed to provide precise information about the level of cognitive functioning at a given point in time, test results must be contrasted with the estimated level of functioning before the injury. The discrepancy between current and premorbid estimates of function is essential, and provides an indication of the degree of deterioration compared with an individually controlled preinjury standard. Traditionally, premorbid ability has been estimated via indirect sources. For example, the vocabulary subtest of the Wechsler adult intelligence scale (WAIS) 1 was widely used as it was assumed to be relatively resistant to the effects of acquired brain damage. However, several recent studies have suggested that vocabulary is not immune to such effects. 2-4 Similarly, intellectual level has been estimated using equations derived from demographic variables-for example, Wilson et al 5 produced a formula based on the WAIS standardisation sample which predicted 54%, 53%, and 42% of the variance in WAIS full scale verbal and performance IQ, respectively. Crawford and Allan 6 recently showed that demographic variables accounted for 54% of WAIS-R variance in a large (n=200) United Kingdom sample. However, other studies have suggested that the percentage of variance explained may be less impressive when using demographic equations alone in clinical samples. 7 8 For example, Perez et al 9 have recently looked at the potential of demographic regression equations to predict premorbid intellectual level in patients with head injury. They reported that the correlation between the equations and WAIS-R verbal IQ wasr =0.52. However, clearly such an approach only accounts for 27% of the variance in observed intelligence, leaving a vast amount of error variance. In attempting to improve on vocabulary and demographically based methods of estimating premorbid intellectual level, Nelson and McKenna 10 suggested that reading ability may provide a more accurate estimate. Johnstoneet al 11 used reading ability from the wide range achievement test-revised (WRAT-R) to estimate premorbid ability in 97 outpatients who had had traumatic brain injury. They found the WRAT-R to be a useful measure of premorbid ability, and highlighted the greatest relative reductions in cognitive flexibility after head injury. Nelson and O'Connel...