Abstract
Background: Impaired self-awareness (ISA) is common among patients with severe traumatic brain injury (TBI) and contributes to poorer functional outcome. There is keen interest in improving the understanding of this disorder as the neuroanatomic substrate of posttraumatic ISA is poorly understood.
Objective: Determine whether (1) greater number of brain lesions, (2) greater volume of right hemisphere lesions, or (3) greater volume of frontal lesions is associated with greater levels of ISA after TBI.
Design: Prospective, observational study.
Participants: Ninety-one TBI admissions to one of 2 National Institute on Disability and Rehabilitation Research TBI Model System (TBIMS) programs. Subjects met TBIMS inclusion criteria plus (1) resolution of posttraumatic amnesia (PTA) prior to rehabilitation discharge and (2) initial postinjury computerized tomography (CT) scan available as a hard copy and as an electronic file.
Methods: CT scan lesions outlined by a board-certified neuroradiologist were measured using NIH Image, and resulting calculated lesion volumes/scan variables compared against demographic characteristics, TBI severity variables, and ISA variables measured by the Awareness Questionnaire (AQ) at the time of PTA resolution.
Results: Most subjects (78%) had at least 1 lesion on emergent CT, and contusion volumes varied in all regions of interest. Patients rated their functioning as more intact on the AQ than ratings of treating clinicians, consistent with ISA. Greater injury severity was associated with a greater degree of ISA. Multivariable linear regression revealed that, after adjustment for other predictors, the number of brain lesions was predictive of degree of ISA. Right hemisphere contusion or frontal lobe contusion volumes, however, were not predictive of degree of ISA.
Conclusions: ISA was significantly associated with the number, but not with location or volume of focal lesions early after TBI. Posttraumatic ISA may reflect disruption in the integrated operation of broadly distributed neural networks, with lesion burden in any specific region being less relevant than disruption across multiple regions. Further imaging research is warranted to confirm these findings and to provide insight into the distributed networks required for self-awareness.