Determining neurological prognosis in patients with severe traumatic brain injury: a survey of Canadian intensivists

Abstract

Introduction: Current prognostic information following severe traumatic brain injury (TBI) is limited. Wide variation in perceptions regarding prognosis determination among clinicians involved in the care of TBI patients likely exists. Following a recent survey of Canadian intensivists (1), we conducted a survey of Canadian neurosurgeons to better understand prognosis determination and decisions concerning the level of care in severe TBI patients. Hypothesis: We believe that a wide variation in practice regarding prognosis determination in this population exists. Methods: We generated items to assess the perceived current utilization and utility of different tests, and to identify factors thought to influence perception of prognosis and level of care determination. We posed direct questions and scenario-based questions with 5 point-Likert scales. We pre-tested the questionnaire for clinical sensibility, content validity and performed test-retest reliability assessment prior to its administration. Canadian neurosurgeons were identified by contacting all level I and II trauma centers. An electronic-based survey was first disseminated followed by a paper questionnaire sent only to non-respondents. Results: The response rate was 54% (93/172). Most respondents worked in teaching hospitals (88%) and were practicing for more than 10 years (54%). Eighty percent considered a poor neurological prognosis at 1 year to correspond to a Glasgow Outcome Scale score of 1, 2 or 3. More than 68% considered that an accurate prediction of prognosis would be most helpful within the first 7 days following TBI. Most respondents identified bedside monitoring (>80%), the clinical exam (>95%), computed tomography scan (>98%) and magnetic resonance imaging (>80%) as being probably or definitely useful for evaluating prognosis, as opposed to somatosensory evoked potentials and electroencephalography (<50%) or brain microdialysis (<15%). Based on a scenario of a 25 year-old man with severe TBI, 29 % of respondent agreed or strongly agreed that this patient had a poor prognosis at 1 year (GOS 1-3), while 44% disagreed or strongly disagreed and 27% had no opinion. In this scenario, 83% of neurosurgeons reported that they were uncomfortable or very uncomfortable to recommend withdrawal of life support measures, 9% were comfortable or very comfortable to do so and 9% had no opinion. When the scenario was modified, comorbidities, trauma severity, diffuse axonal injury, and worsened clinical exam significantly influenced perception of prognosis and level of care determination. Conclusions: There is significant variation in perceptions of neurological prognosis and level of care determination among Canadian neurosurgeons. A better understanding of the factors that influence prognosis perception and determination in TBI patients is required considering the impact/importance of this information on day-to-day care.