Objective: Posttraumatic amnesia is superior to the initial Glasgow Coma Scale score for predicting traumatic brain injury recovery, but it takes days/weeks to assess. Here, we examined whether return of visual fixation-a potential marker of higher cognitive function-within 24 hours of ICU admission could be used as an early predictor of traumatic brain injury recovery.
Design: Two-phase cohort study.
Setting: Level-I trauma ICU.
Patients: Moderate-to-severe traumatic brain injury discharged alive between 2010 and 2013.
Interventions: None.
Measurements and main results: Return of visual fixation was assessed through standard behavioral assessments in 181 traumatic brain injury patients who had lost the ability to fixate at ICU admission (phase 1) and compared with posttraumatic amnesia duration and the initial Glasgow Coma Scale score to predict performance on the Glasgow Outcome Scale-Extended 10-40 months after injury (n = 144; phase 2a). A subgroup also completed a visual attention task (n = 35; phase 2b) and a brain MRI after traumatic brain injury (n = 23; phase 2c). With an area under the curve equal to 0.85, presence/absence of visual fixation at 24 hours of ICU admission was found as performant as posttraumatic amnesia (area under the curve, 0.81; difference between area under the curve, 0.04; p = 0.28) for predicting patients' Glasgow Outcome Scale-Extended score. Conversely, the initial Glasgow Coma Scale score was not (area under the curve, 0.63). Even when controlling for age/medication/CT scan findings, fixation remained a significant predictor of Glasgow Outcome Scale-Extended scores (β, -0.29; p < 0.05). Poorer attention performances and greater regional brain volume deficits were also observed in patients who could not fixate at 24 hours of ICU admission versus those who could.
Conclusions: Visual fixation within 24 hours of ICU admission could be as performant as posttraumatic amnesia for predicting traumatic brain injury recovery, introducing a new variable of interest in traumatic brain injury outcome research.