1. The contribution of psychological factors to recovery after mild traumatic brain injury: Is cluster analysis a useful approach?
- Author
-
Snell, Deborah L., Surgenor, Lois J., Hay-Smith, E Jean C., Williman, Jonathan, and Siegert, Richard J.
- Subjects
- *
ADAPTABILITY (Personality) , *ANALYSIS of variance , *ATTITUDE (Psychology) , *ATTRIBUTION (Social psychology) , *BRAIN injuries , *CHI-squared test , *CLUSTER analysis (Statistics) , *CONFIDENCE intervals , *CONVALESCENCE , *STATISTICAL correlation , *DISEASES , *PSYCHOLOGICAL distress , *HEALTH attitudes , *HOSPITAL emergency services , *LONGITUDINAL method , *EVALUATION of medical care , *MEDICAL rehabilitation , *MENTAL illness , *PATIENTS , *PROBABILITY theory , *PSYCHOLOGICAL tests , *QUESTIONNAIRES , *REHABILITATION centers , *RESEARCH funding , *SCALE analysis (Psychology) , *STATISTICAL hypothesis testing , *STATISTICS , *T-test (Statistics) , *COMORBIDITY , *SOCIOECONOMIC factors , *SYMPTOMS , *EFFECT sizes (Statistics) , *REPEATED measures design , *SEVERITY of illness index , *REHABILITATION for brain injury patients , *POSTCONCUSSION syndrome , *DATA analysis software , *DESCRIPTIVE statistics , *ODDS ratio , *PROGNOSIS ,RESEARCH evaluation - Abstract
Objectives: Outcomes after mild traumatic brain injury (MTBI) vary, with slow or incomplete recovery for a significant minority. This study examines whether groups of cases with shared psychological factors but with different injury outcomes could be identified using cluster analysis. Method: This is a prospective observational study following 147 adults presenting to a hospital-based emergency department or concussion services in Christchurch, New Zealand. This study examined associations between baseline demographic, clinical, psychological variables (distress, injury beliefs and symptom burden) and outcome 6 months later. A two-step approach to cluster analysis was applied (Ward's method to identify clusters, K-means to refine results). Results: Three meaningful clusters emerged (high-adapters, medium-adapters, low-adapters). Baseline cluster-group membership was significantly associated with outcomes over time. High-adapters appeared recovered by 6-weeks and medium-adapters revealed improvements by 6-months. The low-adapters continued to endorse many symptoms, negative recovery expectations and distress, being significantly at risk for poor outcome more than 6-months after injury (OR (good outcome) = 0.12; CI = 0.03-0.53; p < 0.01). Conclusions: Cluster analysis supported the notion that groups could be identified early post-injury based on psychological factors, with group membership associated with differing outcomes over time. Implications for clinical care providers regarding therapy targets and cases that may benefit from different intensities of intervention are discussed. [ABSTRACT FROM AUTHOR]
- Published
- 2015
- Full Text
- View/download PDF