Multidimensional Classification and Prediction of Outcome Following Traumatic Brain Injury.

Objectives: This study aimed to identify outcome clusters among individuals with traumatic brain injury (TBI), 6 months to 10 years post-injury, in an Australian rehabilitation sample, and determine whether scores on 12 dimensions, combined with demographic and injury severity variables, could predict outcome cluster membership 1 to 3 years post-injury.

Setting: Rehabilitation hospital.

Participants: A total of 467 individuals with TBI, aged 17 to 87 (M = 44.2 years), 70% male, with mean post-traumatic amnesia 24 days (range 0.5-455 days), were assessed a mean of 3.4 years post-injury (range 0.5-10 years). A subgroup of 138 participants was also evaluated as rehabilitation inpatients and followed up 1 year post-injury.

Design: Prospective observational study.

Main measures: TBI Quality of Life subscales (Upper Extremity, Pain Interference, Headache Pain, Anxiety, Resilience, Emotional and Behavioral Dyscontrol, General Cognitive Concerns, Independence, and Economic Quality of Life Scale), Neurobehavioral Symptom Inventory, Family Assessment Device General Functioning Scale, Wechsler Adult Intelligence Scale-IV Letter-Number-Sequencing and Coding, Rey Auditory Verbal Learning Test, Trail Making Test Part A, Verbal Fluency Test, Word Memory Test, Participation Assessment with Recombined Tools-Objective, and Glasgow Outcome Scale-Extended.

Results: K-means cluster analysis revealed 5 clusters across 12 dimensions: Good Outcome, High Cognition, Poor Cognition, Poor Outcome, and Poor Adjustment, aligning with Sherer and colleagues' clusters. Inpatient assessments (n = 138) identified profiles predictive of outcome group membership. Participants with Good Outcomes exhibited lower anxiety and higher independence, self-esteem, and resilience, despite some cognitive deficits. High Cognition correlated with robust Economic and Family Support. Poor Cognition aligned with impaired cognitive function but positive psychosocial ratings suggest limited self-awareness. Poor Outcome featured low initial cognitive scores and poor psychosocial adjustment. Poor Adjustment participants, without inpatient cognitive impairments, reported persistent pain, physical symptoms, and emotional distress.

Conclusions: Findings support the evaluation of cognitive and psychosocial factors during rehabilitation to predict outcomes with potential to inform rehabilitative interventions to optimize outcomes.