Factor analysis of an outcome interview for use in clinical trials of traumatically brain-injured patients: a preliminary study.

Abstract

To study the factor structure, internal consistency, concurrent validity, and sensitivity to detect change in patient report of problems of a structured interview in relationship with accepted outcome measures. Outcome status of patients with severe traumatic brain injury participating in a randomized, phase III, multicenter clinical trial was assessed at 6 mo postinjury using the Glasgow Outcome Scale, the Disability Rating Scale, and the Severe Traumatic Brain Injury Outcome Interview. Exploratory factor analysis of the Severe Traumatic Brain Injury Outcome Interview produced a meaningful five-factor model: (1) activities of daily living; (2) cognitive; (3) affective; (4) behavioral; and (5) instrumental activities of daily living. The internal consistency of the factors ranged from moderate (0.61 instrumental activities of daily living) to high (0.94 activities of daily living); the interfactor correlations were moderate. The summed factor scores were significantly correlated with measures of global outcome: the Glasgow Outcome Scale (r = 0.66; P < 0.0001) and the Disability Rating Scale (r = 0.61; P < 0.0001). Patient report of cognitive problems correlated moderately with the neuropsychological tests. The summed factor scores were sensitive to change over time. Overall, the interview assessed the major important features of outcome pertinent to traumatic brain injury and demonstrated greater sensitivity to subtle changes over time than the unidimensional approaches, such as the Glasgow Outcome Scale and Disability Rating Scale.