Autonomic Impairment in Severe Traumatic Brain Injury: A Multimodal Neuromonitoring Study

Objectives: Autonomic impairment after acute traumatic brain injury has been associated independently with both increased morbidity and mortality. Links between autonomic impairment and increased intracranial pressure or impaired cerebral autoregulation have been described as well. However, relationships between autonomic impairment, intracranial pressure, impaired cerebral autoregulation, and outcome remain poorly explored. Using continuous measurements of heart rate variability and baroreflex sensitivity we aimed to test whether autonomic markers are associated with functional outcome and mortality independently of intracranial variables. Further, we aimed to evaluate the relationships between autonomic functions, intracranial pressure, and cerebral autoregulation. Design: Retrospective analysis of a prospective database. Setting: Neurocritical care unit in a university hospital. Subjects: Sedated patients with severe traumatic brain injury. Measurements and Main Results: Waveforms of intracranial pressure and arterial blood pressure, baseline Glasgow Coma Scale and 6 months Glasgow Outcome Scale were recorded. Baroreflex sensitivity was assessed every 10 seconds using a modified cross-correlational method. Frequency domain analyses of heart rate variability were performed automatically every 10 seconds from a moving 300 seconds of the monitoring time window. Mean values of baroreflex sensitivity, heart rate variability, intracranial pressure, arterial blood pressure, cerebral perfusion pressure, and impaired cerebral autoregulation over the entire monitoring period were calculated for each patient. Two hundred and sixty-two patients with a median age of 36 years entered the analysis. The median admission Glasgow Coma Scale was 6, the median Glasgow Outcome Scale was 3, and the mortality at 6 months was 23%. Baroreflex sensitivity (adjusted odds ratio, 0.9; p = 0.02) and relative power of a high frequency band of heart rate variability (adjusted odds ratio, 1.05; p < 0.001) were individually associated with mortality, independently of age, admission Glasgow Coma Scale, intracranial pressure, pressure reactivity index, or cerebral perfusion pressure. Baroreflex sensitivity showed no correlation with intracranial pressure or cerebral perfusion pressure; the correlation with pressure reactivity index was strong in older patients (age, > 60 yr). The relative power of high frequency correlated significantly with intracranial pressure and cerebral perfusion pressure, but not with pressure reactivity index. The relative power of low frequency correlated significantly with pressure reactivity index. Conclusions: Autonomic impairment, as measured by heart rate variability and baroreflex sensitivity, is significantly associated with increased mortality after traumatic brain injury. These effects, though partially interlinked, seem to be independent of age, trauma severity, intracranial pressure, or autoregulatory status, and thus represent a discrete phenomenon in the pathophysiology of traumatic brain injury. Continuous measurements of heart rate variability and baroreflex sensitivity in the neuromonitoring setting of severe traumatic brain injury may carry novel pathophysiological and predictive information.