Using Disability Rating Scale Recovery Curves to Predict PASAT Performance After Closed Head Injury
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Objective: Existing predictive models of cognitive outcome following closed head injury have been largely based on a single time-point. Using archival data, the current study sought to improve upon existing models by predicting cognitive outcome at six months post-injury from a model of the rate of recovery of global functioning over four time-points: hospital discharge, one, three and six months post-injury. Participants and Method: Data from 91 individuals with complicated mild, moderate, and severe closed head injury who had participated in CPHS approved, NIH funded research that involved the collection of global outcome data and neuropsychological testing at six months post injury were used. Disability Rating Scale (DRS) scores from discharge, one, three and six months post-injury were selected along with Paced Auditory Serial Addition Test (PASAT) scores at six months post injury. The PASAT is a task that involves multiple cognitive domains, including processing speed, sustained and divided attention, and working and immediate memory. . Growth curve analysis was used to fit individual growth curves to the DRS scores and then to produce a best-fit recovery curve model. The utility of this model for predicting PASAT scores at six months was determined. Age and severity of injury variables were then added to the model to determine their utility for predicting PASAT scores at six months. Results: Statistical analyses revealed that only the intercept of the DRS recovery curves significantly predicted PASAT performance at six months post-injury. This finding suggested that only the level of DRS score at one month post-injury, and not the trajectory of recovery, was predictive of later PASAT performance. Higher DRS scores, indicated by larger intercepts, were associated with worse PASAT performance. Age was observed to significantly moderate the relationship between the intercept of the DRS recovery curve and PASAT performance at six months post-injury. Other demographic and severity of injury variables were not observed to significantly moderate the relationship between the intercept of the DRS recovery curves and PASAT performance. Conclusion: The change of DRS scores over time was fit best by a quadratic growth curve model with random intercept, linear and curvilinear parameters. Only initial DRS scores were significant predictors of later cognitive performance. Age was the only significant moderator of the relationship between initial DRS scores and PASAT performance. Future research could utilize the current study methodology to evaluate the ability of DRS recovery curves to predict performance on less cognitively demanding neuropsychological tests.