Patterns in withdrawal of life-sustaining treatment in trauma patients with and without moderate to severe traumatic brain injury: an ACS-TQIP retrospective cohort study.

Background: Trauma-specific palliative care guidelines are limited, and physicians rely heavily on experience to guide goals-of-care discussions, leading to substantial practice variation. Existing literature does not adequately account for the marked heterogeneity among trauma patients. How traumatic brain injury influences the timing and determinants of delayed withdrawal of life-sustaining treatment in patients with concomitant major extracranial injuries remains poorly understood. Methods: We conducted a retrospective cohort study of patients ≥16 years with withdrawal of life-sustaining treatment in the American College of Surgeons Trauma Quality Improvement Program database (2021-2023). Patients were stratified by the presence of moderate to severe traumatic brain injury. The primary outcome was time to withdrawal of life-sustaining treatment. Our secondary aim was to assess predictors of delayed withdrawal (>75th percentile for each group) using separate binary logistic regression models. Results: Among 41 990 patients, 23 101 (55%) had a moderate to severe traumatic brain injury. Withdrawal occurred earlier in patients with moderate to severe traumatic brain injury (median 4 days (IQR 2-9) vs. 6 days (3-12), p<0.0001). The probability of delayed withdrawal increased non-linearly with increasing complication burden, plateauing near 100% after seven complications, with the first four having the greatest marginal effect. In patients with moderate to severe traumatic brain injury, unique predictors of delayed withdrawal included diffuse axonal injury, decompressive surgery, and intracranial pressure monitoring. The only major extracranial injury with predictive value was major chest injury (OR 1.3, 95% CI 1.1 to 1.4, p<0.001). Conclusions: Withdrawal of life-sustaining treatment occurs earlier in patients with moderate to severe traumatic brain injury and appears to be driven by distinct factors. Complications may serve as practical triggers for goals-of-care discussions. In polytrauma patients with moderate to severe traumatic brain injury, withdrawal decisions appear to be predominantly influenced by factors related to neurological injury. These findings support pursuing a more nuanced approach to palliative care research in the trauma intensive care unit. Level of evidence: Level III.