Factor structure and measurement invariance of post-concussion symptom ratings on the Health and Behaviour Inventory across time, raters, and groups: An A-CAP study

• Published in: Journal of the International Neuropsychological Society 2023-05, Vol.29 (4), p.346-359
• Main Authors: Zhang, Cherri, Tang, Ken, Zemek, Roger, Beauchamp, Miriam H., Craig, William, Doan, Quynh, Yeates, Keith Owen
• Format: Article
• Language: English
• Subjects: Brain Concussion - complications; Brain Concussion - diagnosis; Brain Concussion - psychology; Child; Children; Children & youth; Cognitive ability; Concussion; Emergency medical care; Factor Analysis, Statistical; Head injuries; Health Behavior; Hospitals; Humans; Parents - psychology; Pediatrics; Post-Concussion Syndrome - diagnosis; Post-Concussion Syndrome - etiology; Post-Concussion Syndrome - psychology; Quality of life; Ratings & rankings; Trauma; Traumatic brain injury
• ISSN: 1355-6177; 1469-7661
• DOI: 10.1017/S1355617722000340

To validate the two-factor structure (i.e., cognitive and somatic) of the Health and Behaviour Inventory (HBI), a widely used post-concussive symptom (PCS) rating scale, through factor analyses using bifactor and correlated factor models and by examining measurement invariance (MI). PCS ratings were obtained from children aged 8-16.99 years, who presented to the emergency department with concussion (n = 565) or orthopedic injury (OI) (n = 289), and their parents, at 10-days, 3-months, and 6-months post-injury. Item-level HBI ratings were analyzed separately for parents and children using exploratory and confirmatory factor analyses (CFAs). Bifactor and correlated models were compared using various fit indices and tested for MI across time post-injury, raters (parent vs. child), and groups (concussion vs. OI). CFAs showed good fit for both a three-factor bifactor model, consisting of a general factor with two subfactors (i.e., cognitive and somatic), and a correlated two-factor model with cognitive and somatic factors, at all time points for both raters. Some results suggested the possibility of a third factor involving fatigue. All models demonstrated strict invariance across raters and time. Group comparisons showed at least strong or strict invariance. The findings support the two symptom dimensions measured by the HBI. The three-factor bifactor model showed the best fit, suggesting that ratings on the HBI also can be captured by a general factor. Both correlated and bifactor models showed substantial MI. The results provide further validation of the HBI, supporting its use in childhood concussion research and clinical practice.