Object: The aim of this study was to examine the variables influencing the mode and location of death in patients admitted to a neurosurgical intensive care unit (NICU), including the participation of a newly appointed neurointensivist (NI).
Methods: Data from all patients admitted to a university hospital NICU were prospectively collected and compared between 2 consecutive 19-month periods before and after the appointment of an NI.
Results: One thousand eighty-seven patients were admitted before and 1279 after the NI's appointment. The withdrawal of life support (WOLS) occurred in 52% of all cases of death. Death following WOLS compared with survival was independently associated with an older patient age (OR 1.04/year, 95% CI 1.03-1.05), a higher University Hospitals Consortium (UHC) expected mortality rate (OR 1.05/%, 95% CI 1.04-1.07), transfer from another hospital (OR 3.7, 95% CI 1.6-8.4) or admission through the emergency department (OR 5.3, 95% CI 2.4-12), admission to the neurosurgery service (OR 7.5, 95% CI 3.2-17.6), and diagnosis of an ischemic stroke (OR 5.4, 95% CI 1.4-20.8) or intracerebral hemorrhage (OR 5.7, 95% CI 1.9-16.7). On discharge from the NICU, 54 patients died on the hospital ward (2.7% mortality rate). A younger patient age (OR 0.94/year, 95% CI 0.92-0.96), higher UHC-expected mortality rate (OR 1.01/%, 95% CI 1-1.03), and admission to the neurosurgery service (OR 9.35, 95% CI 1.83-47.7) were associated with death in the NICU rather than the ward. There was no association between the participation of an NI and WOLS or ward mortality rate.
Conclusions: The mode and location of death in NICU-admitted patients did not change after the appointment of an NI. Factors other than the participation of an NI-including patient age and the severity and type of neurological injury-play a significant role in the decision to withdraw life support in the NICU or dying in-hospital after discharge from the NICU.