The high-level mobility assessment tool (HiMAT) for traumatic brain injury. Part 2: content validity and discriminability

• Published in: Brain injury Vol. 19; no. 10; p. 833
• Main Authors: Williams, G P, Robertson, V, Greenwood, K M, Goldie, P A, Morris, M E
• Format: Journal Article
• Language: English
• Published: England 01.09.2005
• Subjects: Adult; Brain Injuries - diagnosis; Brain Injuries - rehabilitation; Female; Gait; Humans; Male; Movement Disorders - diagnosis; Outcome Assessment (Health Care) - methods; Reproducibility of Results; Sensitivity and Specificity
• ISSN: 0269-9052
• DOI: 10.1080/02699050500058711

(i) To assess the measurement properties of the high-level mobility assessment tool (HiMAT) for people with traumatic brain injury (TBI), (ii) to measure the extent to which the HiMAT is a uni-dimensional, discriminative hierarchical outcome scale. The content validity was assessed using a three-stage process of investigating internal consistency, factor analysis and Rasch analysis. The uni-dimensionality of the HiMAT items was also tested. Discriminability was investigated by correlating raw and logit scores obtained from Rasch analysis. The study was conducted at a major rehabilitation facility using a convenience sample of 103 adults with TBI. The internal consistency for the high-level items was very high (Cronbach's alpha = 0.99). Principal axis factoring identified several balance items as belonging to a second factor not related to high-level mobility, hence these items were excluded. Rasch analysis identified several misfitting items, such as walking around a figure of eight and stopping from a run, which were also excluded. Logit scores were used to exclude clustered and, therefore, redundant items. Raw scores correlated very highly (r = 0.98) with logit scores, indicating that raw scores provided good discriminability and were suitable for use by clinicians. The HiMAT, which assesses higher-level mobility requirements of people with TBI for return to pre-accident social, leisure and sporting activities, is a uni-dimensional and discriminative scale for quantifying therapy outcomes.