Traumatic brain injury is one of the most common acquired disabilities in the United States, and there is significant heterogeneity with this population in terms of cognitive, behavioral, and psychological characteristics. The aim of this study is to determine how well certain neuropsychological tests, premorbid IQ estimates and psychosocial variables are able to discriminate between mild and moderate/severe traumatic brain injury (TBI). The study proposed three hypotheses: (1) neuropsychological tests alone will be able to discriminate mild from moderate/severe TBI. (2) Neuropsychological tests, in addition to premorbid IQ estimates, will be able to better discriminate mild from moderate/severe TBI than using neuropsychological tests alone. (3) Neuropsychological tests, when combined with premorbid IQ estimates and psychosocial factors will provide the best degree of discrimination between mild and moderate/severe TBI.

The results of the study indicated that Hypothesis 1 was partially supported—only two of the neuropsychological measures, Wisconsin Card Sorting Test Perseveration Errors and Booklet Category Test demonstrated a very modest ability to discriminate between mild and moderate/severe TBI. A single discriminant function was calculated for the two neuropsychological measures with 58.3% of the sample being correctly classified using these two neuropsychological variables. Hypothesis 2 combined premorbid IQ estimates and neuropsychological test scores and yielded one discriminant function. A total of 60.0% of the sample was correctly classified using the Verbal IQ and Performance IQ Estimated WTAR scores along with the previous neuropsychological variables. Adding psychosocial variables to neuropsychological test scores and premorbid IQ estimates yielded one discriminant function for Hypothesis 3. A weak relationship was found between predictor variables and ACRM Criteria outcomes with a slightly higher overall classification rate of 65.8% being achieved. A secondary analysis was performed using the presence of positive brain MRI as the single predictor variable to distinguish the TBI groups. This discriminant analysis yielded an overall classification rate of 72.5% suggesting that this single variable is a more powerful predictor of group membership than neuropsychological or psychological variables in this sample of patients. In sum, neuropsychological test scores, premorbid IQ estimates, and psychosocial measures possess only a mild ability to discriminate mild from moderate/severe TBI populations.

These results are discussed in terms of the potential limitations of neuropsychological and psychological test findings in characterizing those that sustain mild versus moderate TBI. The study also suggests that for the specific purpose of distinguishing these groups, the brain MRI may be a better predictor.