Alternative Structure Models of the Traumatic Brain Injury Rehabilitation Needs Survey: A Veterans Affairs TBI Model Systems Study

Abstract

ObjectiveTo explore the factor structure of the Rehabilitation Needs Survey (RNS). DesignSecondary analysis of observational cohort study who were 5-years post-traumatic brain injury (TBI). SettingFive Inpatient Rehabilitation Facilities. ParticipantsVeterans enrolled in the TBI Model Systems longitudinal study who completed the RNS at 5-year follow-up (N=378). Main Outcome Measure(s)RNS. ResultsRNS factor structure was examined with exploratory factor analysis (EFA) with oblique rotation. Analyses returned 2- and 3-factor solutions with Cronbach alphas ranging from 0.715 to 0.905 and corrected item-total correlations that ranged from 0.279 to 0.732. The 2-factor solution accounted for 61.7% of the variance with ≥3 exclusively loading items on each factor with acceptable internal consistency metrics and was selected as the most parsimonious and clinically applicable model. Ad hoc analysis found the RNS structure per the EFA corresponded with elements of the International Classification of Functioning, Disability and Health (ICF) conceptual framework. All factors had adequate internal consistency (α≥0.70) and 20 of the 21 demonstrated good discrimination (corrected item-total correlations≥0.40). ConclusionsThe 2-factor solution of the RNS appears to be a useful model for enhancing its clinical interpretability. Although there were cross-loading items, they refer to complex rehabilitation needs that are likely influenced by multiple factors. Alternatively, there are items that may require alteration and redundant items that should be considered for elimination.