Background: No standards exist regarding decision making for comatose patients, especially concerning life-saving treatments. The aim of this retrospective, single-center study was to analyze outcomes and the decision-making process at the end of life (EOL) in patients with traumatic brain injury (TBI) in a Swiss academic tertiary care hospital.
Methods: Consecutive admissions to the surgical intensive care unit (ICU) with stays of at least 48 hours between January 1, 2012 and June 30, 2015 in patients with moderate to severe TBI and with fatality within 6 months after trauma were included. Descriptive statistics were used.
Results: Of 994 ICU admissions with TBI in the study period, 182 had an initial Glasgow Coma Scale <13 and a length of stay in the ICU >48 hours. For 174 of them, a 6-month outcome assessment based on the Glasgow Outcome Scale (GOS) was available: 43.1% (36.0%-50.5%) had favorable outcomes (GOS 4 or 5), 28.7% (22.5%-35.9%) a severe disability (GOS 3), 0.6% (0%-3.2%) a vegetative state (GOS 2), and 27.6% (21.5%-34.7%) died (GOS 1). Among the GOS 1 individuals, 45 patients had a complete dataset (73% men; median age, 67 years; interquartile range, 43-79 years). Life-prolonging therapies were limited in 95.6% (85.2%-99.2%) of the cases after interdisciplinary prognostication and involvement of the surrogate decision maker (SDM) to respect the patient's documented or presumed will. In 97.7% (87.9%-99.9%) of the cases, a next of kin was the SDM and was involved in the EOL decision and process in 100% (96.3%-100.0%) of the cases. Written advance directives (ADs) were available for 14.0% (6.6%-27.3%) of the patients, and 34.9% (22.4%-49.8%) of the patients had shared their EOL will with relatives before trauma. In the other cases, each patient's presumed will was acknowledged after a meeting with the SDM and was binding for the EOL decision.
Conclusions: At our institution, the majority of deaths after TBI follow a decision to limit life-prolonging therapies. The frequency of patients in vegetative state 6 months after TBI is lower than expected; this could be due to the high prevalence of limitation of life-prolonging therapies. EOL decision making follows a standardized process, based on patients' will documented in the ADs or on preferences assumed by the SDM. The prevalence of ADs was low and should be encouraged.