Using confirmatory factor analysis to identify symptom clusters in individuals with acquired disabilities
摘要
Acquired disabilities, such as spinal cord injury, traumatic brain injury, limb loss (or other major limb injury/illness), or stroke, are typically accompanied by an array of physical and emotional symptoms and social challenges that interfere with recovery and affect quality of life (QOL). Despite advances in personalized medicine, interrelationships between symptoms remain unclear, and clinicians often treat individual symptoms in isolation. The objective of this study was to identify clusters of symptoms that occur across sudden-onset disability populations, with a long-term goal of identifying targets for clinical intervention.
MethodsParticipants were 755 adults who sustained a spinal cord injury, traumatic brain injury, stroke, or sudden-onset limb-threatening injury/illness. Participants completed a battery of physical, emotional, and social patient reported outcomes measures and performance-based cognitive tests at baseline and after 14–18 months. Confirmatory factor analysis (CFA) of 26 indicator variables was used to reduce dimensions and identify clusters of symptoms.
ResultsCFA results provide support for 10 factors: Cognition, Economic QOL, Negative Affect, Positive Affect, Psychological Adjustment, Physical Function, Independence, Sleep Impairment/Fatigue, Pain, and Social Health. Global model fit was acceptable: χ2 (255) = 875.0, RMSEA = 0.059, CFI = 0.948. The three emotional health clusters were highly correlated. The 10 factors demonstrated stability over time.
ConclusionThese results can inform the development of composite scores and/or symptom indices to identify and compare broad areas of functioning and provide a method to detect specific areas of clinical concern. This improved understanding of symptom interrelationships may facilitate individualized care based on specific patient characteristics and, ultimately, lead to improved clinical outcomes for individuals with acquired disabilities.